Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
This action is responsive to communications field on 08/11/2026.
Claims 1-20 are currently pending.
Claim 1 and 15 are independent claims.
Priority
Receipt is acknowledged of certified copies of papers required by 37 CFR 1.55.
Prior Arts
Listed herein below are the prior art references relied upon in this office action:
Humpal et al. (US 12,075,769 B2, which has a priority date of 02/10/2021), referred to as Humpal herein.
Loukili et al. (US 11,690,368 B2, which has a priority date of 01/13/2022), referred to as Loukili herein.
Slaughter et al. (US 10,993,430 B2, which has a priority date of 10/10/2018), referred to as Slaughter herein.
Redden et al. (US 11,093745 B2, which has a priority date of 035/09/2018), referred to as Redden herein.
Walther et al. (US 11,874,367 B2, which has a priority date of 03/04/2019), referred to as Walther herein.
Pickett et al. (US 11,310,954 B2, which has a priority date of 09/12/2018), referred to as Pickett herein.
Hammer et al. (US 10,485,229 B2, which has a priority date of 05/19/2015), referred to as Hammer herein.
King et al. (US 10,949,974 B2, which has a priority date of 04/25/2019), referred to as King herein.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim(s) 1, 3-4, 11, 13, 15, 17 is/are rejected under 35 U.S.C. 103 as being unpatentable over Humpal in view of Loukili further in view of Slaughter.
Regarding Claim 1, Humpal teaches a system mounted in a vehicle, (“Agricultural machine 100 is depicted as an agricultural sprayer that has an operator compartment 102, supported by a frame structure 104, which also supports ground engaging elements 106. In the example shown in FIG. 1A, ground engaging elements 106 are wheels, but they could be tracks or other implementations.” (Col 4, line 34-40), meaning agricultural machine 100 is embodied as an agricultural sprayer supported by a frame structure and traveling on ground engaging wheels or tracks. Additionally, the sprayer’s operator compartment, boom, nozzles, and sensors are all carried by this same frame and ground engaging structure);
the system comprising: a boom arrangement that comprises a predefined number of electronically controllable sprayer nozzles (“Spray system 108 also illustratively includes a boom structure 118 that supports a plurality of controllable nozzle bodies 120. Nozzle bodies 120 can include (as shown in more detail below) an electronic controller that receives commands over a network, such as a controller area network-CAN, or other data communication protocols. The nozzle body 120 can also include one or more controllable valves that can be moved between an open position and a closed position.” (Col 4, line 55-63), meaning the spray system includes a boom structure 118 supporting a plurality of controllable nozzle bodies 120, each equipped with an electronic controller capable of receiving commands over a network such as a CAN bus. Because the nozzle bodies are electronically addressable and individually actuatable through this controller and valve arrangement, Humpal discloses a boom arrangement comprising a predefined number of electronically controllable sprayer nozzles);
and a plurality of image-capture devices configured to capture a plurality of field-of-views (FOVs) of a plurality of defined areas of an agricultural field, wherein each of the plurality of image-capture devices is configured to capture an image frame corresponding to a field-of-view (FOV) of a defined area of the agricultural field; (“the image sensors 122 are disposed across boom 118 so that their fields of view cover all of the area of field 112 forward of nozzle bodies 120, as agricultural machine 100 travels through the field.” (Col 5, line 21-24) and “Image sensors 122 may be optical sensors which capture images by sensing radiation in the optical spectrum” (Col 5, line 14-16), meaning the image sensors 122 are optical sensors positioned across the boom, and each sensor’s field of view covers a defined area of the field ahead of the nozzles. Each sensor therefore captures an image frame corresponding to the FOV of a defined area of the agricultural field. This satisfies the conditional requirement of the claim);
and one or more hardware processors configured to: (“Each image processing module 124 can have its own set of processors 224, data store 226, communication system 228, camera calibration system 230, image remapping system 232, confidence level generator 234, image stitching system 236, white balance correction system 238, row identification system 240, target identification processor 242, nozzle identification system 244 (which can include lateral adjustment determination component 245 and other items 247), nozzle activation control system 256, and output generator 248.” (Col 10, line 21-30), disclosing that each processing module is provided with its own set of processors 224, which, together with the confidence level generator and target identification processor relied upon carry out the image analysis, target identification, and nozzle control functions. This satisfies the conditional requirement of the claim);
obtain a plurality of images corresponding to the plurality of FOVs from the plurality of image-capture devices, wherein each image of the plurality of images corresponds to the image frame captured by a corresponding image-capture device of the plurality of image-capture devices; (“The image sensors 122 are illustratively coupled to one or more image processing modules 124. The image processing modules 124 illustratively process the images captured by image sensors 122 to identify targets (e.g., weeds 116 or rows 114) on field 112 over which agricultural machine 100 is traveling.” (Col 5, line 25-30) and “weed locator 388 applies the mapping coefficients (or calibration transforms) corresponding to this particular image sensor at this particular boom height or boom location, identified by mapping coefficient identifier 340 (FIG. 5).” (Col 32, line 10-14), meaning the image processing modules process the images captured by the image sensors, and each image is processed using the coefficients that correspond to the particular image sensor that captured it. Each image therefore corresponds to the image frame captured by a corresponding image capture device);
Humpal does not teach receiving geospatial location correction data from an external device placed at a fixed location and coordinates associated with the boom arrangement mounted in the vehicle.
However, Loukili teaches receive geospatial location correction data from an external device placed at a fixed location in the agricultural field and geospatial location coordinates associated with the boom arrangement mounted on the vehicle, (“Position sensor(s) 232 are configured to determine a geographic position of machine 202 on the field, and can include, but are not limited to, a Global Navigation Satellite System (GNSS) receiver that receives signals from a GNSS satellite transmitter. It can also include a Real-Time Kinematic (RTK) component that is configured to enhance the precision of position data derived from the GNSS signal.” (Col 4, line 40-47), meaning the position sensor 232 determines the position of the machine using a GNSS receiver and an RTK component that improves the precision of that position. As is well known in the art, RTK positioning uses a reference station at a known, fixed location to generate correction data and transmit it to the receiver on the moving machine. Loukili therefore discloses receiving geospatial location correction data from an external device at a fixed location and resulting location coordinates for the machine carrying the boom. This satisfies the conditional requirement of the claim);
Humpal and Loukili do not teach the geospatial location correction data provides a geolocation of each image frame corresponding to the plurality of images.
However, Slaughter teaches wherein the geospatial location correction data provides a geolocation of each image frame corresponding to the plurality of images; (“Coincident with image acquisition step 22, location information may optionally be retrieved at step 24 to assign location data to the acquired image.” (Col 5, line 31-33) and “The geolocation of the plant 32 is determined, using geolocation sensor 48 (FIG. 3) (e.g. an onboard RTK-GPS or odometry sensor), at the time each plant 32 is inspected for the presence of the crop signal.” (Col 10, line 62-65), meaning assigning location data, read from its geolocation sensor when the image is captured, to each acquired image, and that sensor can be an onboard RTK GPS. The RTK corrected position therefore serves as the geolocation of the image frame. In the combination, the correction data received in Loukili provides the geolocation of each image frame);
Humpal teaches execute mapping of pixel data of weeds or a crop plant in an image frame corresponding to one of the plurality of images to distance information from a reference position of the boom arrangement when the vehicle is in motion, (“Identifying the weed location locates the pixel coordinates of the center of the pixel clusters (that represent the weeds) in relation to the ROI for the image sensor. In one example, weed locator 388 applies the mapping coefficients (or calibration transforms) corresponding to this particular image sensor at this particular boom height or boom location, identified by mapping coefficient identifier 340 (FIG. 5).” (Col 32, line 7-22)” and “Weed locator 388 then determines, spatially, where the weed is, geographically, in the field.” (Col 32, line 4-5), meaning it converts the pixel location of a detected weed into a location relative to the boom while the machine is in motion, and then determines where the weed is geographically in the field. This satisfies the conditional requirement of the claim);
Humpal and Loukili do not teach to determine a geolocation of the weeds or the crop plant based on pixel data corresponding to the weeds or the crop plant in the image frame and the geolocation of the image frame.
However, Slaughter teaches wherein, to execute the mapping, the one or more hardware processors are further configured to determine a geolocation of the weeds or the crop plant based on pixel data corresponding to the weeds or the crop plant in the image frame and the geolocation of the image frame; (“As the plant care assembly 100 travels along the row, a continuous panoramic view of the row is created by acquiring a sequence of images, processing them as described above in real-time to detect the crop plants and recording their geolocation using the centroid information within each image for the crop plus the odometry geolocation for the image itself.” (Col 13, line 61-66), meaning determines a plant’s geolocation from the plant’s centroid position within the image and the geolocation of the image itself. The centroid is the pixel data of the plant in the image frame, and the geolocation of the image itself is the geolocation of the image frame. This satisfies the conditional requirement of the claim);
Humpal teaches and cause a specific set of electronically controllable sprayer nozzles from amongst the predefined number of electronically controllable sprayer nozzles to operate based on a defined confidence threshold and the executed mapping of pixel data, (“Adjusting the sensitivity value through an operator interface mechanism is indicated by block 646. The sensitivity value corresponds to a minimum pixel cluster size (or pixel cluster) that will qualify as a target identification and will thus trigger a spray command. Providing the sensitivity level to correspond to a minimum pixel cluster size is indicated by block 648.” (Col 31, line 55-61) and “Filter 384 then filters targets (e.g., weeds) detected in the image by size, based upon the sensitivity value.” (Col 31, line 64-65), disclosing a sensitivity value, adjustable by the operator through the operator interface, that corresponds to a minimum pixel cluster size a detected object must meet before it qualify as a target identification and trigger a spray command, with Filter 384 applying this value to filter detected targets by size before a spray command issues. Therefore, a detected pixel cluster must satisfy this defined sensitivity value before a corresponding nozzle is caused to operate and that determination follows directly from the pixel level mapping. This satisfies the conditional requirement of the claim);
Humpal teaches wherein the defined confidence threshold is indicative of a detection sensitivity of the crop plant, (“Adjusting the sensitivity value through an operator interface mechanism is indicated by block 646. The sensitivity value corresponds to a minimum pixel cluster size (or pixel cluster) that will qualify as a target identification and will thus trigger a spray command.” (Col 31, line 55-59), meaning that the same sensitivity value disclosed above directly governs whether a detected pixel cluster qualifies as a target identification. Humpal refers to this quantity as a sensitivity value and because it is a value (rather than any other system parameter) that determines whether a given detection is treated as a valid target, this discloses the defined confidence threshold is indicative of a detection sensitivity of the crop plant);
Humpal teaches and wherein a change in the defined confidence threshold causes a corresponding change in operation of the predefined number of electronically controllable sprayer nozzles. (“The sensitivity value corresponds to a minimum pixel cluster size (or pixel cluster) that will qualify as a target identification and will thus trigger a spray command.” (Col 31, line 56-59), meaning the sensitivity value determines whether a given pixel cluster satisfies the minimum size needed to qualify as a target and thereby triggers a spray command, any change to that value necessarily changes which detected objects to qualify, and therefore changes which nozzles are triggered to operate. This satisfies the conditional requirement of the claim).
At the time of the invention, it would have been obvious to a person of ordinary skill in art to combine Humpal’s agricultural sprayer, which has boom mounted image sensors and a weed locator that determines weed locations from pixel data, with Loukili’s RTK enhanced position sensor that receives correction data from a fixed reference station, and with the Slaughter’s teaching of assigning location data to each acquired image and determining a plant’s geolocation from its position in the image and the geolocation of the image. As modified, Humpal’s system would receive RTK correction data from an external fixed reference station as taught by Loukili, use the corrected position as the geolocation of each captured image frame as taught by Slaughter, and determine the geolocation of each weed or crop plant from its pixel data in that frame and the geolocation of the frame as taught by Slaughter.
The motivation for doing so would have been to improve the accuracy of the weed and crop plant locations that Humpal determines. Loukili states that its RTK component is “configured to enhance the precision of position data derived from the GNSS signal” (Col 4, line 45-47), and Slaughter shows that the position associated with an image can be combined with the position of a plant within that image to record the plant’s location. A more precise image geolocation directly produces a more precise weed location, which allows Humpal to activate the correct nozzle at the correct time. Because all three references are directed to locating plants in an agricultural field for treatment, and the combination applies known techniques to obtain a more accurate location, a person of ordinary skill in the art would have expected predictable results.
Regarding Claim 3, Humpal teaches the system according to claim 1, wherein the one or more hardware processors are configured to update the defined confidence threshold in response to a change in a quality parameter of the captured plurality of FOVs of the plurality of defined areas of the agricultural field. (“If the marker is no longer detectable due to dust or other obscurants, the confidence level is reduced. Nozzle speed can also be used to calculate or infer image blur or other poor image quality. Further, boom height can affect image quality as well. If the boom is too high, image resolution may suffer. If the boom is too low, the ROI may not be sufficiently large. Row identification confidence level detector 348 generates a confidence level indicative of how confident the system is that it has adequately identified crop rows in the image” (Col 33, line 57-67) and “The image quality and the various confidence levels generated by confidence level generator 234 may be considered by other logic or components in the system in determining whether and how to apply material to the field.” (Col 14, line 51-52), meaning the system monitors multiple image quality parameters, specifically, the presence of dust or obscurants that render markers undetectable, nozzle speed that causes image blur, and boom height that affects both image resolution and the size of the region of interest captured by each sensor and when any of these quality parameters change, the confidence level generated by confidence level generator 234 is correspondingly reduced. Furthermore, these confidence levels generated in response to image quality are then considered by other logic or components in the system in determining whether and how to apply materials to the fields. The confidence threshold governing the materials application decision responds directly to changes in image quality parameters (adjusting upward or downward as dust, blur, and boom height conditions change).
Regarding Claim 4, Humpal teaches the system according to claim 1, wherein the specific set of electronically controllable sprayer nozzles are operated further based on a predefined operating zone of the vehicle, wherein the predefined operating zone defines a range of speed of the vehicle in which an accuracy of the detection sensitivity of the crop plant is greater than a threshold. (“When the image sensor 122 is traveling very quickly, it can be more difficult to identify targets and activate the nozzle 120 to spray targets quickly enough that the target can be covered by the material sprayed by the nozzle (because the image sensor is moving so quickly over the ground).” (Col 14, line 23-28), “Spray operation confidence level detector 350 generates a confidence level indicative of how confident the system is that it can adequately deliver material to an identified target. (Col 34, line 17-19) and “In one example, it may be that the nozzle is moving at an excessive rate of speed.” (Col 34, line 23-24), meaning when Humpal’s image sensor travels very quickly, such as during a turn, it becomes more difficult to identify targets and to activate the corresponding nozzle quickly enough for the sprayed material to reach that target, and that Humpal’s spray operation confidence level detector 350 generates a confident level reflecting how confident the system is that it can adequately deliver material to an identified target, a confidence level Humpal expressly identifies as being affected by, among other things, the nozzle moving at an excessive rate of speed. Because it is thus disclosed that machine speed and detection accuracy are directly related and that accuracy falls below an acceptable level once speed becomes excessive, the existence of a range of vehicle speeds within which detection and spray accuracy remain acceptable. It would have been obvious to define and enforce such a speed range as an operating parameter of Humpal’s system, establishing a predefined operating zone bounded by the speeds at which Humpal’s own confidence level remains above an acceptable threshold, in order to ensure reliable target identification and nozzle activation during normal field operation, which was itself the express objective underlying Humpal’s spray operation confidence level detector).
Regarding Claim 11, Humpal teaches the system according to claim 1, wherein the one or more hardware processors are further configured to communicate control signals to operate a plurality of different sets of electronically controlled sprayer nozzles at different time instants during a spray session. (“Nozzle identification system 244 then associates the weed (or other target) location with one or more nozzles on boom 118. Associating the target location with the nozzles is indicated by block 664.” (Col 32, line 23-25) and “Delay time on generator 406 then generates the delay time which will be time elapsed prior to activating the identified nozzle.” (Col 40, line 52-54), meaning as the machine moves across the field, each newly identified target is associated with whichever nozzle(s) on boom 118 corresponds to that target’s location, different targets trigger different nozzle sets over the course of a single pass. For each such identified nozzle, the delay time on generator computes a distinct elapsed time value before that specific nozzle is activated, timed to when the target reaches it. Because target location (and therefore the associated nozzle sets) continuously changes as the machine travels, and each nozzle set is activated only at its own computed delay time, control signals are communicated to operate different sets of nozzles at different time instants throughout the spray session).
Regarding Claim 13, Humpal teaches the system according to claim 1, wherein the one or more hardware processors are further configured to: distinguish between two different green looking objects corresponding to crop plants and weeds when a first defined confidence threshold is set; or distinguish between a type of crop plant and a type of weed when a second defined confidence threshold is set different from the first defined confidence threshold. (“the sensitivity value corresponds to a minimum pixel cluster size (or pixel cluster) that will qualify as a target identification and will thus trigger a spray command…. Filter 384 then filters targets (e.g., weeds) detected in the image by size, based upon the sensitivity value. This type of filtering is indicated by block 652 in the flow diagram of FIG. 22B. The weeds (or other targets) are identified by finding green (or segmented) blobs (pixel clusters) in the image of the sufficient size that they meet the sensitivity valve.” (Col 31 line 56 – Col 32, line 3), meaning ------the target identification system identifies weeds and other targets by searching the captured image for green, or segmented, pixel clusters (green looking blobs) and only those pixel clusters that satisfy the defined sensitivity value qualify as target identifications and triggers a spray command, with Filter 384 performing this filtering based upon the sensitivity value. The system identifies targets by finding green pixel clusters in the image and filtering them through the sensitivity value, therefore the system inherently distinguishes between different green looking objects (green pixel clusters that are large enough to meet the sensitivity value threshold) and those that are not. This corresponds to the distinction between two different green looking objects corresponding to crop plants and weeds when a first defined confidence threshold is set. Additionally, because the sensitivity value is operator adjustable, setting it to a different value allows the same filter logic to perform finer, type level discrimination between specific types of crop plants and types of weeds, corresponding to the claimed second defined confidence threshold set differently from the first. This satisfies the conditional requirement of the claim).
Regarding Claim 15, a method claim that incorporates the system of claim 1, is being rejected under same rationale as claim 1.
Regarding Claim 17, it is being rejected under same rationale as claim 3.
Claim(s) 2, 16 is/are rejected under 35 U.S.C. 103 as being unpatentable over Humpal, Loukili, Slaughter further in view of Redden.
Regarding Claim 2, Humpal, Loukili, and Slaughter do not teach defined confidence threshold set in real-time or near real-time in an artificial intelligence (AI) model of the system or pre-set in the AI model via user interface (UI) rendered on a display device communicatively couple to one or more hardware processors.
However, Redden teaches the system according to claim 1, wherein the defined confidence threshold is set in real-time or near real-time in an artificial intelligence (Al) model of the system or pre-set in the Al model via a user interface (UI) rendered on a display device communicatively coupled to the one or more hardware processors. (“The plant detection model can be generated using a variety of machine-learning tools including, but not limited to, neural networks, support vector machines, and decision trees. In one embodiment, the plant detection model is a modified version of the Single Shot MultiBox Detector (SSD) neural network.” (Col 2, line 4-9) and “The output may also include a numerical confidence of the plant detection model 180 in its prediction regarding the bounding box. Together, as will be described in Section III below the bounding boxes, and in some implementations the numerical confidences as well, are used to determine an action taken by the platform 100.” (Col 8, line 21-27), disclosing an agricultural sprayer platform equipped with a trained artificial intelligence model. A modified SSD convolutional neural network that processes camera images of the field in real time as the machine travels and generates a numerical confidence value for each detected plant bounding box, which confidence value is then used directly to determine what treatment action the machine takes. This teaches AI model that produces, operates on and uses confidence values within its real-time detection pipeline to control spray decision, corresponding to the claimed confidence threshold operating in an AI model of the system to govern nozzle activation).
At the of the invention, it would have been obvious to a person of ordinary skill in the art to combine the agricultural boom sprayer architecture of Humpal (which discloses a confidence threshold governing nozzle activation based on pixel cluster size detection), Loukili’s RTK based geospatial positioning, and Slaughter’s teaching of assigning location data to each acquired image and determining a plant’s geolocation from its position in the image with the artificial intelligence approach of Redden which discloses a trained neural network AL model that generates numerical confidence values for each plant detection on the same type of moving agricultural platform.
The motivation for doing so would have been that Humpal’s own system already identifies a need for a confidence based plant detection threshold to determine which detection qualify as targets and trigger spray commands, and Redden expressly teaches that a neural network AI model performing precisely this function of generating per-detection confidence values used to determine treatment actions, achieves substantially improved accuracy and adaptability over conventional rule based detection approaches under the same varying field conditions. Substituting Redden’s AI neural network detector for Humpal’s pixel cluster rule-based detector, while Humpal’s existing confidence threshold framework and operator interface for adjusting that threshold, is a predictable application of a well-known, superior technique to obtain the expected result of improved detection performance. Because Humpal, Loukili, and Redden are directed to the same problem of real-time, image-based crop and weed detection on a moving agricultural spray platform and because the confidence threshold Humpal already employs to gate nozzle activation maps directly onto the role of the confidence value Redden’s AI model generates to determine treatment decisions, the combination requires no more than ordinary skill to implement and yields predictable results.
Regarding Claim 16, it is being rejected under same rationale as claim 2.
Claim(s) 5-6, 8-10, 18-19 is/are rejected under 35 U.S.C. 103 as being unpatentable over Humpal, Loukili, Slaughter further in view of Walther.
Regarding Claim 5, Humpal, Loukili, and Slaughter do not teach determining a height of a tallest crop plant and set a boom height from the ground plane based on the determined height of the tallest crop plant.
However, Walther teaches determine a height of a tallest crop plant from among a plurality of crop plants from a ground plane in the agricultural field; (“The ascertainment of the plant height taking place with the aid of a signal of a radar sensor” (Abstract) and “The reflections of plant objects 22 result from a height range, which extends from the ground up to a maximum growth height of the field crop. Correspondingly, the reflections which are presumably close to a maximum growth height in the field crop are advantageously more heavily weighted in the ascertainment of the plant object curve.” (Col 4, line 25-31), meaning automatically determining plant height using radar measurement of plants in the field. Furthermore, “maximum growth height” of the detected vegetation corresponds to determining the height of the tallest crop plant among multiple crop plants. Thus, identifying the tallest plant within the sensed region);
and set a boom height from the ground plane based on the determined height of the tallest crop plant. (“the plant height of the field crops to be harvested represents an input variable for operating the agricultural machine…. The plant height of the field crops also represents an input variable for setting a treatment device, for example a spraying device.” (Col 1, line 7-18), meaning the determined plant height is used as an input variable for setting a spraying device. Because the spraying boom is a principal adjustable component of a spraying device whose operating height determines spray application relative to the crop, the measured plant height is used to set the operating height of the spraying boom so that an appropriate clearance is maintained above the crop. This satisfies the conditional requirement of the claim).
At the time of the invention, it would have been obvious to a person of ordinary skill in art to modify the agricultural spraying system of Humpal, Loukili, and Slaughter’s RTK based geospatial positioning to incorporate Walther’s plant height determination techniques, including determination of maximum crop height, and to use the determined crop height as an operating parameter for the spraying system.
The motivation for doing so would have been to improve control of the spraying operation based on actual crop conditions. Walther expressly teaches “The plant height of the field crops also represents an input variable for setting a treatment device, for example a spraying device.” (Col 1, line 16-18). Because all references are directed to agricultural machine operation and crop treatment, a person of ordinary skill in the art would have been motivated to apply Walther’s crop height information within the spraying system of Humpal and Loukili in order to set the operating height of the spray boom relative to the crop canopy. Doing so would predictably improve spray placement accuracy, maintain desired boom to crop clearance, and reduce the risk of crop damage, which is recognized goals in the field of agricultural spraying.
Regarding Claim 6, Humpal teaches the system according to claim 5, wherein the one or more hardware processors are further configured to determine an upcoming time slot to spray a chemical based on the executed mapping of the pixel data, the defined confidence threshold, and the boom height set from the ground plane. (“Delay time on generator 406 then generates the delay time which will be time elapsed prior to activating the identified nozzle.” (Col 40, line 52-54), “The variables may also include the nozzle and boom position which indicates the height of the boom above the ground, and the orientation of the nozzle given that boom position...... Nozzle activation control system 246 also receives the target (e.g., weed) location 872, from weed locator 388 in target identification processor 242” (Col 40, line 40-53), meaning Humpal’s delay time on generator 406 computes the delay time that will elapse before the identified nozzle is activated, using variables that include the boom’s height above that ground and the orientation of the nozzle at that boom position, together with the target location received from weed locator 388. This delay time computation determines when the identified nozzle should be activated based on both the boom height established in Claim 5 and the pixel mapped target location established in Claim 1, discloses determining an upcoming time slot to spray a chemical based on the executed mapping of the pixel data and the boom height set from the ground plane. Additionally, “Filter 384 then filters targets (e.g., weeds) detected in the image by size, based upon the sensitivity value. This type of 65 filtering is indicated by block 652 in the flow diagram of FIG. 22B.... Weed locator 388 then determines, spatially, where the weed is, geographically, in the field.” (Col 31 line 64 – Col 31, line 5), meaning Filter 384 filters detected targets by size according to the sensitivity value before weed locator 388 determines the qualifying target’s spatial, geographic location in the field. A target’s location is only passed forward to the delay time and nozzle activation determination once it has satisfied this sensitivity value, disclosing that the determination of the upcoming time slot to spray is also based on the defined confidence threshold. This satisfies the conditional requirement of the claim).
Regarding Claim 8, Loukili teaches the system according to claim 6, wherein the one or more hardware processors are further configured to determine one or more regions in the agricultural field where to spray the chemical based on the executed mapping of pixel data and the defined confidence threshold. (“The computing system also includes weed identification logic configured to identify locations of weed plants in the field based on the identification of the first and second image portions” (Abstract), meaning the computing system includes weed identification logic configured to identify the geographic locations of weed plants in the field based on the identification of first and second image portions, that is based on a comparison of pixel level image data. This logic identifies specific field location requiring treatment directly from mapped image data. Additionally, Humpal teaches “Adjusting the sensitivity value through an operator interface mechanism is indicated by block 646. The sensitivity value corresponds to a minimum pixel cluster size (or pixel cluster) that will qualify as a target identification and will thus trigger a spray command.” (Col 31, line 55-59), meaning that the sensitivity value already established in Claim 1, the claimed confidence threshold determines which detected pixel clusters qualify as targets in the first place, Loukili’s weed identification logic maps that qualifying detection to a specific geographic region for treatment Therefore, it would have been obvious to combine Loukili’s region identification teaching with Humpal’s confidence threshold based qualification logic because both operate on the same underlying image and pixel data such that a detection must first satisfy Humpal’s confidence threshold before Loukili’s logic identifies where that qualifying detection should be treated. This satisfies the conditional requirement of the claim).
Regarding Claim 9, Humpal teaches the system according to claim 8, wherein the specific set of electronically controllable sprayer nozzles from amongst the predefined number of electronically controllable sprayer nozzles are caused to operate specifically at the determined one or more regions in the agricultural field for a first time slot that corresponds to determined upcoming time slot. (“Nozzle identification system 244 then associates the weed (or other target) location with one or more nozzles on boom 118. Associating the target location with the nozzles is indicated by block 664.” (Col 32, line 23-25) and “Delay time on generator 406 then generates the delay time which will be time elapsed prior to activating the identified nozzle.” (Col 40, line 52-54), meaning nozzle identification system 244 first pinpoints exactly which nozzle(s) on boom 118 sit at the determined region where the target was found, that is the specific set of nozzles corresponding to that region, per block 664. Delay time on generator 406 then computes the precise elapsed time before that identified nozzle is actually activated, timed so the nozzle fires as the machine’s travel brings the nozzle into position over the target. That compound delay time on value is the “first time slot” at which the nozzle set operates, and it necessarily corresponds to the “upcoming time slot” already determined for that region in Claim 6, the delay time is calculated specifically to align nozzle activation with the previously identified upcoming spray opportunity. This satisfies the conditional requirement of the claim).
Regarding Claim 10, Humpal teaches the system according to claim 9, wherein the one or more hardware processors are further configured to control an amount of spray of the chemical for the first time slot from each of the specific set of electronically controllable sprayer nozzles by regulating an extent of opening of a valve associated with each of the specific set of electronically controllable sprayer nozzles. (“the valves are variable between the on and off positions, such as proportional values. In other examples, a variable flow rate can be achieved through the valves by controlling the pump or by controlling the valves in a pulse width modulated manner (varying the cycle time) or in other intermittent ways” (Col 5, line 6-11) and “Nozzle/valve controller 170 can then control the rate of flow of material through the nozzle bodies 120 based upon the spray rate received from spray rate controller 222.” (Col 10, line 9-12), meaning the values are not simple binary on/off devices, they can be held at intermediate, proportional positions between fully open and fully closed, and the nozzle/valve controller 170 actively drives that variable positioning (whether by modulating the valve itself, pulse width modulating its duty cycle or adjusting the pump) to set the actual flow rate through each nozzle body. The “extend of opening” of the valve is therefore the mechanism by which the “amount of spray” is controlled for each nozzle in the identified set).
Regarding Claim 18, it is being rejected under same rationale as claim 5.
Regarding Claim 19, it is being rejected under same rationale as claim 6.
Claim(s) 7, 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Humpal, Loukili, Slaughter, and Walther further in view of Pickett.
Regarding Claim 7, Humpal, Loukili, Slaughter, and Walther do not teach determining of the upcoming time slot to spray the chemical based on a size of the crop plant occupied in a two-dimensional space in x and y coordinate direction.
However, Pickett teaches determining of the upcoming time slot to spray the chemical is further based on a size of the crop plant occupied in a two-dimensional space in x and y coordinate direction. (“the spray configurator 316 can determine how long it will be until a spray target is positioned under one of the nozzles 118 of the tractor 100 based on information from the spray target determiner 326 and the speed determiner 310, and thereby determine a start time for the spray operation. Similarly, the spray configurator 316 can determine a spray duration based on the size of the spray target using data from the spray target determiner 326, and based on the speed of the tractor 100 from the speed determiner 310.” (Col 22, line 47-56), meaning the spray configurator 316 determines two distinct timing values for each detected spray target. First, a start time for the spray operation, determined from the spray target determiner 326 and the speed determiner 310 based on when the spray target will be positioned under a nozzle and second, a spray duration (how long the nozzle remains active) determined from the measured size of the spray target as provided by the spray target determiner 326, together with the machine’s speed from speed determiner 310. The spray duration, which defines the length of the upcoming spray time slot, is computed directly from the measured size of the spray target and a target’s size is a two-dimensional quantity expressed in both the direction of travel (y-direction) and across the boom width (x-direction)).
At the time of the invention, it would have been obvious to a person of ordinary skill in art to combine Humpal’s boom mounted image detection and nozzle activation timing system, Loukili’s RTK based geospatial positioning, Slaughter’s teaching of assigning location data to each acquired image and determining a plant’s geolocation from its position in the image and Walther’s plant height determination techniques with Pickett’s teaching of calculating spray duration from the measured two-dimensional size of the detected spray target. Specifically, it would have been obvious to incorporate Pickett’s size-based duration calculation into Humpal’s existing nozzle activation control system 246, so that in addition to using target location and machine speed to determine when to begin spraying, Humpal’s system would also use the pixel cluster size of the detected target as determined by Humpal’s target identification processor 242 to determine how long each nozzle should remain active during the spray time slot.
The motivation for doing so would have been to more accurately and completely cover each detected target with chemical during a single pass of the boom sprayer. Humpal’s timing system determines when to activate a nozzle based on target location and speed but does not expressly account for the spatial extent of the target in calculating how long the nozzle remains on. Pickett identifies this problem and teaches that spray duration must be proportional to the target’s size to ensure full coverage, a straightforward and predictable improvement directly addressing the risk of under treating larger targets or over treating areas beyond the target’s boundary.
Regarding Claim 20, it is being rejected under same rationale as claim 7.
Claim(s) 12 is/are rejected under 35 U.S.C. 103 as being unpatentable over Humpal, Loukili, Slaughter further in view of Hammer.
Regarding Claim 12, Humpal teaches configured to receive a user input, via a user interface rendered on a display device, (“When the user interface mechanisms 154 include a display screen, operator input mechanisms can be provided on the display screen. Such operator input mechanisms can include buttons, links, icons, or other user actuatable elements that can be actuated using a point and click device, a touch gesture, a voice input, or other interactions.” (Col 7, line 36-41), thus receiving a user input through a user interface rendered on a display device, wherein an operator may interact with display rendered controls and provide inputs to the system through user actuatable interface elements);
Humpal, Loukili, and Slaughter do not teach the user input corresponding to a user-directed disablement or enablement and automatic activation and deactivation of electronically controllable nozzle during a spray session.
However, Hammer teaches wherein the user input corresponds to a user-directed disablement, or an enablement of one or more electronically controllable nozzles to override an automatic activation and deactivation of the one or more electronically controllable nozzles during a spray session. (“Another alternative example method is to enable or disable individual nozzles 40 within a section 30 to which the scroller has reached (“instant section”). Scrolling through sections 30 is designated as a section adjust, and scrolling through individual nozzles 40 is a nozzle adjust.” (Col 5, line 56-61) and “The center section 32 (or 32A) may be ignored if it is already enabled (e.g. by manual override, another control button or software GUI).” (Col 5, line 20-23), meaning that an operator provides user input to selectively enable or disable individual electronically controllable nozzles during an active spray operation. Additionally, the manual override may supersede the existing operating state of a spray section or nozzle, indicating that the operator directed enablement or disablement overrides the otherwise automatic activation or deactivation of the nozzles during a spray session. This satisfies the conditional requirement of the claim).
At the time of the invention, it would have been obvious to a person of ordinary skill in art to modify Humpal’s boom mounted image detection and nozzle activation timing system, Loukili’s RTK based geospatial positioning and Slaughter’s teaching of assigning location data to each acquired image and determining a plant’s geolocation from its position in the image to incorporate the operator selectable nozzle enable and nozzle disable functionality taught by Hammer. Thereby, allowing a user, through the display-based user interface of Humpal, to selectively enable or disable one or more electronically controllable nozzles and override the automatic nozzle activation and deactivation performed by the spraying system during a spray session.
The motivation for doing so would have been to provide the operator with greater control over spray application in circumstances where automated spraying decisions may require manual intervention. Hammer expressly recognizes the benefit of allowing an operator to selectively engage or disengage individual nozzles, while Humpal and Loukili provide an automated vision-based operator interface. A person of ordinary skill in the art would have understood that incorporating Hammer’s manual nozzle override capability into the automated spraying system of Humpal and Loukili would improve operational flexibility by allowing an operator to immediately override automatic nozzle activation and deactivation when encountering field conditions that warrant human judgement, such as field boundaries, vegetation, obstacles, or other application specific circumstance. Such modification would merely involve the predictable use of a known nozzle control feature within a similar agricultural spraying system to obtain the expected benefit of enhanced operator control and improved spray management.
Claim(s) 14 is/are rejected under 35 U.S.C. 103 as being unpatentable over Humpal, Loukili, Slaughter further in view of King.
Regarding Claim 14, Humpal, Loukili, and Slaughter do not teach to set a third defined confidence threshold to distinguish between a diseased or a non- diseased crop plant and further distinguish weeds from the diseased or the non-diseased crop plants.
However, King teaches set a third defined confidence threshold to distinguish between a diseased or a non- diseased crop plant and further distinguish weeds from the diseased or the non-diseased crop plants, (“In response to detecting the at least one difference, additional images of the plants are requested from the one or more image capture devices to detect the presence of plant disease.” (Abstract) and “At block 718, the classifier may classify the discrete objects into plant and non-plant objects. Upon obtaining a discrete object list for the buffered image, the objects may be sorted by size, from the largest to the smallest. A predetermined threshold may be specified for the maximum size and the minimum size. In this way, only the image objects of a certain size within the predetermined threshold may be processed to reduce the processing load…. After calculation of the rank values, the top X values, X being a predetermined threshold, will be deemed to be representing a plant in the image.” (Col 19, line 25-38), meaning the classifier at block 718 classifies discrete image objects into plant and non-plant categories, sorts them by size from largest to smallest, applies a predetermined threshold for both the maximum and minimum object size to limit which objects are processed, and then uses a second predetermined threshold (the top X ranked values) to determine which remaining objects qualify as representing a plant in the image. Because King is directed to automated plant disease detection, and because this threshold based classification pipeline distinguishes plant objects from non-plant objects and uses a predetermined threshold value to determine which classified plant objects are further processed for disease identification, King discloses a defined, predetermined confidence threshold used to distinguish between diseased and non-diseased crop plants and to further distinguish plant objects from non-plant objects in the image. Thus, teaching the third defined confidence threshold required to distinguish between a diseased or non-diseased crop plant and further distinguish weeds from the diseased or non-diseases crop plants);
Humpal teaches wherein the third defined confidence threshold is different from the first defined confidence threshold and the second defined confidence threshold; or set a fourth defined confidence threshold to further distinguish between a discoloured plant or a non-discoloured plant, identity a growth state of crop plants while additionally distinguishing the crop plants from the weeds. (“Filter 384 then filters targets (e.g., weeds) detected in the image by size, based upon the sensitivity value. This type of filtering is indicated by block 652 in the flow diagram of FIG. 22B. The weeds (or other targets) are identified by finding green (or segmented) blobs (pixel clusters) in the image of the sufficient size that they meet the sensitivity valve.” (Col 31 line 60 – Col 32, line 3), meaning the Filter 384 filters all detected targets (weeds and other objects) by comparing each green pixel cluster against the sensitivity value, with only those clusters of sufficient size that meet the sensitivity value qualifying for further processing and ultimately triggering a spray command. Because sensitivity value is adjustable, a third threshold value set differently from the first and second sensitivity value settings previously established in Claim 13 would (under the same filter logic) govern a distinct category of target discrimination. Distinguishing plants exhibiting abnormal color or size profiles (discolored or growth state differentiated plants) from those that do not, while continuing to distinguish all qualifying plant objects from weeds. A fourth threshold is set to a different value correspondingly enables discrimination between discolored and non-discolored plants or between plants at different growth states and the weeds present in the same field. This satisfies the conditional requirement of the claim).
At the time of the invention, it would have been obvious to a person of ordinary skill in art to combine the plant disease detection and threshold-based classification system of King with the image-based weed and crop detection system of Humpal, Loukili, and Slaughter. Adding King’s disease detection threshold classification logic as a third, distinct threshold operating in Humpal’s existing image processing pipeline, so that after Humpal’s system identifies which green pixel clusters qualify as plants under the first and second confidence threshold setting, King’s classifier and predetermined threshold are then applied to determine whether the qualifying plant objects exhibit characteristics associated with disease, thereby distinguishing diseased from non-diseased crop plants and further distinguishing weed objects from both categories of crop plant.
The motivation for doing so would have been to extend Humpal’s precision spraying capability from basic weed to crop discrimination to health status-based plant discrimination, enabling the sprayer to apply targeted chemical treatment differently depending on whether a crop plant is diseased or healthy. Because different plant diseases may require different chemical interventions or spray rates, and because Humpal’s system already captures the pre plant image data needed to perform disease classification, incorporating King’s disease detection threshold represents a predictable application of a known techniques, image-based disease classification using a predetermined threshold, to achieve a well-recognized and valuable precision agriculture benefit.
Response to Arguments
Applicant's arguments submitted on 08/11/2026 have been fully considered but they are not persuasive.
The applicant appears to argue that Humpal does not receive geospatial location correction data, and that Loukili corrects only the position of the machine. These arguments address the references individually, but the rejection is based on the combination of the references. The rejection already acknowledged that Humpal does not teach receiving the correction data and relied on Loukili for it.
The applicant further argues that Humpal is only an optical calibration tool and does not determine where a weed is located on the ground. Humpal states that the weed locator “determines, spatially, where the weed is, geographically, in the field” (Col 32, line 4-5), and Humpal’s machine includes a geographic position sensor (Col 16, line 10-15). Applicant also argues that Loukili does not apply position data to a captured image. Loukili’s “location system 282 can output GPS coordinates that are associated with the captured image to obtain geo-referenced images 286 that are provided to plant evaluation system 226.” (Col 5, line 67 - Col 6, line 3) and “weed identification logic 338 generates geographic coordinates for each separate field area that has been identified as containing weeds, and stores this data as weed data 340.” (Col 10, line 18-21).
The applicant also relies on features that the claims do not recite, including satellite-based data, a GNSS receiver or RTK data, a precision of about 2.5 cm, a “single-step operation,” and independence from boom-height calibration or machine position data. Claims are given their broadest reasonable interpretation in light of the specification, but limitations from the specification are not read into the claims. Claim 1 still recites mapping pixel data to distance information from a reference position of the boom arrangement, so it does not exclude Humpal’s calibration teachings. Lastly, Applicant did not inspect the Examiner’s statement that RTK positioning uses a reference station at a known, fixed location to provide correction data, so that statement is taken as admitted prior art.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. See PTO-892.
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Contact
Any inquiry concerning this communication or earlier communications from the examiner should be directed to CHANDNI PATEL whose telephone number is (571)272-9661. The examiner can normally be reached Monday-Friday 7am-4pm, every other Friday off.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Scott Baderman can be reached at (571)272-3644. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/CHANDNI PATEL/Examiner, Art Unit 2118
/SCOTT T BADERMAN/Supervisory Patent Examiner, Art Unit 2118