DETAILED ACTION
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 .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's RCE submission filed on 03/31/2026 has been entered.
Response to Amendment
The amendment filed 02/26/2026 has been entered. Claims 1-4, 6-11, 13-17, and 19-20 remain pending in the application, with claims 5, 12, and 18 having been cancelled.
Response to Arguments
Applicant's arguments on pg. 7-8 of the Remarks, filed 02/26/2026, have been fully considered but they are not persuasive. Applicant argues that Derbisz fails to teach amended claims 1, 9, and 14. Examiner respectfully disagrees. Applicant argues that Derbisz’s distance estimation and correction process is “object-centric” and therefore, Derbisz cannot teach “inferring contextual features that characterize individual regions of the image itself and selecting correction factors on a per-region basis based on those region-characterizing contextual features”. The location where the object is located is a region in the image. Thus, inferring contextual features and selecting correction factors on a per-region basis is taught by Derbisz’s method/system. Indicating a plurality of individual regions with bounding boxes ultimately produces the result Applicant argues Derbisz does not teach: correcting predicted distances on a region-by-region basis. There are no limitations recited in amended claims 1, 9, and 14 that explicitly exclude an interpretation where a region in the image is equated with a location of an individual object or its bounding box. In view of the foregoing, Derbisz is maintained to reject the amended independent claims with further explanation below.
Applicant's arguments on pg. 9 of the Remarks, filed 02/26/2026, have been fully considered but they are not persuasive. Examiner respectfully disagrees for the same reasons as those outlined above with respect to the amended independent claims.
Claim Interpretation
Regarding claim 4, the claim interpretation previously set forth in the Final Office Action of 02/09/2026 is withdrawn due to para 93 of the Specification: “The phrase “at least one of ... and ....” as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items. As an example, the phrase “at least one of A, B, and C” includes A only, B only, C only, or any combination thereof (e.g., AB, AC, BC or ABC)”. Thus, the at least one contextual feature in claim 4 is interpreted in accordance with para 93.
Claim Rejections - 35 USC § 102
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claims 1, 9, and 14 are rejected under 35 U.S.C. 102(a)(1) and (a)(2) as being anticipated by Derbisz (U.S. Patent No. 2022/0270371 A1).
Regarding claim 1, Derbisz teaches a system (Derbisz, para 78: “computer system 600”; see Figure 6) comprising:
a processor (Derbisz, para 78: “processor 602”);
a memory in communication with the processor (Derbisz, para 78: “memory 604” of Figure 6) and having a control module, the control module having instructions that, when executed by the processor (Derbisz, Computer program, para 79: “The non-transitory data storage 606 may store a computer program, including the instructions that may be transferred to the memory 604 and then executed by the processor 602”), cause the processor to:
determine predicted distances to objects using an image acquired by a camera of a vehicle (Derbisz, para 24: “the present disclosure is directed at a vehicle comprising the computer system as described herein, a camera configured to acquire the image, and the distance sensor”), wherein the image comprises a plurality of regions (Derbisz, see Figure 1 wherein the image comprises a plurality of different regions characterized by bounding boxes and their objects, para 55: “FIG. 1 shows an illustration 100 of an image and corresponding bounding boxes (for example bounding box 102). Each bounding box may be related to a class of the object determined in the bounding box”; para 9: “This system may detect many object classes at once and may provide their localization”; see also object of Figure 4 steps 402-406 and para 58, para 58: “At 406, a coarse estimation of the distance may be determined based on a distance sensor.”);
infer, for each individual region of the plurality of regions, at least one contextual feature characterizing the individual region of the plurality of regions (Derbisz, Class of the objects, para 58: “At 404, a class of the object may be determined based on the image”; class is inferred for each region, para 55: “Each bounding box may be related to a class of the object determined in the bounding box”), wherein the at least one contextual feature is associated with a correction factor (Derbisz, correction for each class, para 68-69: “According to various embodiments, the distance may be determined based on a hash table for the class of the object. According to various embodiments, the hash table may include a correction (for each class) for determining the distance based on the coarse estimation.”);
correct the predicted distances for objects located in each individual region using the correction factor selected based on the at least one contextual feature characterizing the individual region (Derbisz, correction of each object distance where the objects are related to the plurality of regions defined above, para 58: “At 408, the distance of the object may be determined based on the coarse estimation and based on the class of the object.”; correction is based on the class of each region, para 42: “a further correction may be provided by compensation of distances for objects, which may be implied by choosing an appropriate training strategy which relies on specifying objects positions (heading direction) and sizes (width, length) encapsulating them into training class types, making the classes more granular”), wherein the correction factor differs among the plurality of regions (Derbisz, correction differs for different regions containing different classes, para 69: “the hash table may include a correction (for each class) for determining the distance based on the coarse estimation”; class is also based on region of image wherein object is located, which applies to the plurality of regions in FIG. 1, for example; para 13: “the plurality of orientations comprises left, and/or upper-left, and/or upper, and/or upper right, and/or right, and/or bottom-right, and/or bottom, and/or bottom-left”; para 51: “for example, the class of an object may be “sedan_bottom_right” (e.g., a type of the object of class “sedan”, and an orientation of the object of “bottom right”)”); and
control a movement of the vehicle based on the corrected predicted distances (Derbisz, para 9: “A real-time object detection system which may be applied in a self-driving car may be provided”; see also para 34).
Regarding claim 9, Derbisz teaches a non-transitory computer-readable medium including instructions that, when executed by a processor, cause the processor (Derbisz, para 79: “The non-transitory data storage 606 may store a computer program, including the instructions that may be transferred to the memory 604 and then executed by the processor 602”). All further claim limitations are met by Derbisz because the steps executed by the processor of claim 9 are the same as that of claim 1.
Regarding claim 14, all claim limitations are met by Derbisz because the method steps of claim 14 are the same as that of claim 1.
Claim Rejections - 35 USC § 103
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.
Claims 2-3, 10-11, and 15-16 are rejected under 35 U.S.C. 103 as being unpatentable over Derbisz in view of Tomatsu (U.S. Patent No. 2022/0110511 A1).
Regarding claim 2 (dependent on claim 1), Derbisz teaches wherein the control module further includes instructions that, when executed by the processor, cause the processor to: responsive to determining the at least one contextual feature does not correspond to a correction factor, determine a new correction factor associated with the at least one contextual feature (Derbisz, para 14: “For example, a method for correction vectors construction from training classes assignment may be provided. The object classes may be semantically related, and/or size (for example, width and height) related, and/or orientation specific (for example with granularity: bottom, bottom-left, left, upper-left, up, upper-right, right, bottom-right).”), but fails to explicitly teach determine a difference between one of the predicted distances and an actual distance to at least one object; and determine a new correction factor associated with the at least one contextual feature based, at least in part, on the difference, wherein the new correction factor replaces the correction factor.
However, Tomatsu teaches a system for predicting the distance to an object (Tomatsu, abstract: “detecting a distance to a target object”) including determine a difference between the predicted distance and an actual distance to the at least one object (Tomatsu, para 155: “the creation unit 505F may obtain a difference between the optical distance of the design value and the actual optical distance”); and determine a new correction factor, at least in part, on the difference (Tomatsu, correction information, para 155: “may create the correction information 504F for correcting the optical distances L1, L2, L3, L4, and L5 stored in advance. For example, the creation unit 505F may obtain a difference between the optical distance of the design value and the actual optical distance, and may create the correction information 504F for changing the optical distance of the design value”), wherein the new correction factor replaces the correction factor (Creating new correction information will replace the use of a non-corresponding correction factor, see also note below regarding the combination of Derbisz in view of Tomatsu.). It would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to have combined the correction factor based on the difference, as taught above by Tomatsu, with the system of Derbisz in order to improve the accuracy of the predicted distance (Tomatsu, para 154: “ As a result, even if an error occurs between the optical distance stored in advance and the actual optical distance, the endoscope system 1 can calculate the distance D6 with higher accuracy”). A person of ordinary skill in the art would be able to utilize the above teachings of Tomatsu to add/replace correction factors in the hash table taught by Derbisz (Derbisz, para 68-69).
Regarding claim 3 (dependent on claim 2), Derbisz in view of Tomatsu teach wherein the control module further includes instructions that, when executed by the processor, cause the processor to determine the new correction factor further include instructions that, when executed by the processor, cause the processor to determine the new correction factor responsive to determining the difference satisfies a distance threshold (Tomatsu, para 155: “may create the correction information 504F for changing the optical distance of the design value in a case where the difference is equal to or larger than the threshold value”) that is based, at least in part, on a minimal difference between the one of the predicted distances and the actual distance (Tomatsu, the set threshold indicates the minimum difference value that can be exceeded).
Regarding claim 10, all claim limitations are met and rendered obvious by Derbisz in view of Tomatsu because the steps executed by the processor of claim 10 are the same as that of claim 2.
Regarding claim 11, all claim limitations are met and rendered obvious by Derbisz in view of Tomatsu because the steps executed by the processor of claim 11 are the same as that of claim 3.
Regarding claim 15, all claim limitations are met and rendered obvious by Derbisz in view of Tomatsu because the method steps of claim 15 are the same as that of claim 2.
Regarding claim 16, all claim limitations are met and rendered obvious by Derbisz in view of Tomatsu because the method steps of claim 16 are the same as that of claim 3.
Claims 4 and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Derbisz in view of Wang et al. (CN Patent No. 115115687 A), hereinafter Wang, Mehra et al. (U.S. Patent No. 2023/0099598 A1), hereinafter Mehra, and Chen (CN Patent No. 114339607 B).
Regarding claim 4 (dependent on claim 1), Derbisz teaches wherein the at least one contextual feature includes at least one of: external properties of the vehicle, wherein the external properties include at least one of: a classification of at least one object (Derbisz, Class of the object, para 58: “At 404, a class of the object may be determined based on the image”), but fails to teach wherein the at least one contextual feature includes at least one of: internal properties of the vehicle and external properties of the vehicle, wherein the internal properties include at least one of: dimensions of the vehicle, an operating parameter of the vehicle, and a behavior of a driver of the vehicle, and wherein the external properties include at least one of: a classification of the at least one object and a weather condition.
Wang teaches a similar system (Wang, pg. 1, 2nd to last paragraph: “Determine the lane line depth information of the target pixel based on the height information, wherein the lane line depth information is used to represent the relative horizontal positional relationship between the lane line and the target vehicle”) including at least one contextual feature including an internal property of the vehicle, wherein the internal property includes dimensions of the vehicle (Wang, vehicle height, pg. 9, 3rd to last paragraph: “information h of the target vehicle”). The dimensions of the vehicle are associated with a correction factor (Wang, pg. 9, 3rd paragraph: “lane line depth information is corrected according to a correction factor”) that is utilized to correct the predicted distance of an object (Wang, pg. 9, 3rd paragraph: “Due to the angle between the monocular camera and the target vehicle, there is a deviation in the shooting angle of the monocular camera. Therefore, when it is determined that there is a deviation in the shooting angle of the monocular camera, the target pixel is corrected by a correction factor.”). It would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to have combined the use of the dimensions of the vehicle, as taught by Wang, in the system of Derbisz in order to improve distance prediction accuracy by accounting for the size of the vehicle of which the camera is attached (Wang, pg. 6, 4th paragraph: “if the monocular camera is set at different positions of the target vehicle, the height information of the monocular camera will also be different”; see also previous citation from pg. 9). Omitting the dimensions of the vehicle could result in an erroneous predicted distance based on where the camera is located on the vehicle.
Additionally, Mehra teaches a similar system (Mehra, abstract) including at least one contextual feature including internal properties of the vehicle, wherein the internal properties include an operating parameter of the vehicle (Mehra, zones analyzed by the computer depend on the gear, an operating parameter, of the vehicle, para 13: “The second zone may be forward of the first zone relative to the vehicle when a gear of the vehicle is a forward gear, and the second zone may be rearward of the first zone when the gear is a reverse gear”; different zones around the vehicle correspond to preset distances from the vehicle, para 44: “The zones 106 have preset boundaries relative to the vehicle 100, i.e., the zones 106 are stored in the memory of the computer 102, and when the vehicle 100 moves, the zones 106 move with the vehicle 100”; para 46: “The computer 102 can determine which zone 106 includes a current or predicted position of an object 104. For example, the computer 102 can compare the coordinates of the current or predicted position with the boundaries of the zones 106 stored in memory”) and a behavior of a driver of the vehicle (Mehra, putting the car in forward versus reverse indicates a behavior of the driver and will result in similar changes to that of the operating parameter described above, para 32: “The human operator may control the propulsion 108 via, e.g., an accelerator pedal and/or a gear-shift lever”). The operating parameter of the vehicle and a behavior of a driver of the vehicle are associated with factors utilized to analyze the distance of an object (Mehra, see para 46 citation above, coordinates of the position are compared with the boundaries of the zone; see the relationship between zones and distance from the vehicle in Figure 6). It would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to have combined the use of an operating parameter of the vehicle and a behavior of a driver of the vehicle, as taught by Mehra, in the system of Derbisz in view of Wang in order to improve the accuracy of the distance prediction by accounting for whether the vehicle is in forward or reverse (Mehra, see paragraph 13 citation above). For example, the distance between the coordinates of objects in a zone and the boundaries of that zone in relation to the vehicle may be different when the vehicle is in reverse versus forward gear (gear position affects which zone is utilized in determining the position of the object).
Further, Chen teaches a similar system (Chen, abstract: “determining a target adjustment coefficient between the first UWB module and the second UWB module; and adjusting the first distance according to the target adjusting coefficient to obtain a second distance. By adopting the embodiment of the application, accurate distance measurement can be realized”) including at least one contextual feature including an external property of the vehicle, wherein the external property includes a weather condition (Chen, pg. 10, last paragraph: “Wherein the external environmental parameter may be at least one of: the environmental temperature, the environmental humidity, the atmospheric pressure, the interference intensity of the external magnetic field, the longitude and latitude, the wind speed, the weather, and the like are not limited herein”). The weather condition is an environmental parameter associated with a correction factor (Chen, target adjustment coefficient, pg. 10, 2nd to last paragraph: “determining the target adjusting coefficient corresponding to the target external environment parameter according to the mapping relation between the preset external environment parameter and the adjusting coefficient”) that is utilized to correct the predicted distance of an object (Chen, pg. 11, last paragraph: “adjusting the first distance according to the target adjusting coefficient to obtain a second distance”). Further, Chen teaches the utilization of multiple types of properties to influence the target adjusting coefficient (Chen, pg. 11, 7th paragraph: “carrying out weighting operation according to the first adjusting coefficient, the second adjusting coefficient, the first weight and the second weight to obtain the target adjusting coefficient”). It would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to have combined the use of the weather condition and combination of multiple types of properties, as taught by Chen, in the imaging system of Derbisz in view of Wang and Mehra in order to obtain a more accurate distance measurement by accounting for multiple internal/external factors that may affect the system’s distance prediction methods (Chen, abstract: “By adopting the embodiment of the application, accurate distance measurement can be realized”). Although implemented in a UWB system, a person of ordinary skill in the art would recognize the benefit of implementing external environmental parameters and combining multiple types of properties when predicting distance in the system of Derbisz, similar to the way that object classification influences the corrected distance in Derbisz.
Regarding claim 17, all claim limitations are met and rendered obvious by Derbisz in view of Wang, Mehra, and Chen because the steps executed by the processor of claim 17 are the same as that of claim 4.
Claims 6, 13, and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Derbisz in view of Zhu et al. (Zhu, J., & Fang, Y. (2019). Learning object-specific distance from a monocular image. In Proceedings of the IEEE/CVF International Conference on computer vision (pp. 3839-3848).), hereinafter Zhu.
Regarding claim 6 (dependent on claim 1), Derbisz teaches wherein the control module further includes instructions that, when executed by the processor, cause the processor to determine the at least one contextual feature is associated with the correction factor further include instructions that correlate the at least one contextual feature with the correction factor (Derbisz, associated correction vectors are based on object class, para 14: “For example, a method for correction vectors construction from training classes assignment may be provided. The object classes may be semantically related, and/or size (for example, width and height) related, and/or orientation specific (for example with granularity: bottom, bottom-left, left, upper-left, up, upper-right, right, bottom-right).”), but fails to explicitly teach when executed by the processor, cause the processor to train a machine learning model to correlate the at least one contextual feature with the correction factor.
However, Zhu teaches an object-specific distance estimation method (Zhu, abstract and Figure 1 on pg. 3839) that, when executed by the processor, cause the processor to train a machine learning model (Zhu, Figure 3 on pg. 3842) to correlate the at least one contextual feature with the correction factor (Zhu, pg. 3842, paragraph before section 3.2: “Implying a prior knowledge of the correlation between the object class and its real size and shape, the classifier encourages our model to learn features that can be leveraged in estimating more accurate distances”; see loss equations 1-3 on pg. 3842 and further explanation as follows). Zhu teaches a machine learning model that is trained to correlate a contextual feature (Zhu, class, size, shape of objects) with a correction factor (Zhu, loss value, which is based on the object class, learns features of the size and shape of objects, and is used to predict a more accurate distance value). It would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to have combined the machine learning model of Zhu with the system of Derbisz in order to improve the accuracy of the association between contextual features and correction factors by using a trained model that learns how to minimize distance errors based on real object information, such a class and corresponding size and shape (Zhu, pg. 3842 citation and loss value equations cited above).
Regarding claim 13, all claim limitations are met and rendered obvious by Derbisz in view of Zhu because the steps executed by the processor of claim 13 are the same as that of claim 6.
Regarding claim 19, all claim limitations are met and rendered obvious by Derbisz in view of Zhu because the method steps of claim 19 are the same as that of claim 6.
Claims 7 and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Derbisz in view of Ye et al. (U.S. Patent No. 2025/0005777 A1), hereinafter Ye.
Regarding claim 7 (dependent on claim 1), Derbisz fails to explicitly teach wherein the control module further includes instructions that, when executed by the processor, cause the processor to present the corrected predicted distances to a driver of the vehicle.
However, Ye teaches a similar system (Ye, abstract: “device includes a processor that obtains first data and second data from different cameras of a vehicle. The processor generates a depth map based on a disparity between the first data and the second data”) that presents the corrected predicted distances to a driver of the vehicle (Ye, para 23: “The display 160 may potentially include additional features such as object detection or distance estimation to further enhance driver awareness and safety”). It would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to have combined the presentation of distance information to the driver of a vehicle, as taught by Ye, with the system and corrected predicted distances of Derbisz in order to enhance driver awareness and safety while operating the vehicle (Ye, see last citation).
Regarding claim 20, all claim limitations are met and rendered obvious by Derbisz in view of Ye because the method steps of claim 20 are the same as that of claim 7.
Claim 8 is rejected under 35 U.S.C. 103 as being unpatentable over Derbisz in view of Chawla et al. (Chawla, J., Thakurdesai, N., Godase, A., Reza, M., Crandall, D., & Jung, S. H. (2021, September). Error diagnosis of deep monocular depth estimation models. In 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (pp. 5344-5649). IEEE.), hereinafter Chawla.
Regarding claim 8 (dependent on claim 1), Derbisz fails to explicitly teach wherein the correction factor is at least one of: an overestimation factor or an underestimation factor, wherein the overestimation factor is associated with the predicted distances being greater than an actual distance to at least one object, and wherein the underestimation factor is associated with the predicted distances being less than an actual distance to the at least one object.
However, Chawla teaches a system for correction depth estimation to an object with a correction factor (Chawla, amount distance is corrected by in the Depth Error Correction Network, pg. 5322, Method section: “We also introduce a technique for correcting the detected errors (Sec III-B)”; see also abstract), wherein the correction factor is at least one of: an overestimation factor or an underestimation factor, wherein the overestimation factor is associated with the predicted distances being greater than an actual distance to at least one object, and wherein the underestimation factor is associated with the predicted distances being less than an actual distance to the at least one object (Chawla, pg. 5324, section B in the left column: “One application of DEDN is to refine the depth map by trying to correct the errors it has identified. We can do this by incrementally increasing depths of pixels that are predicted to be under-estimated, and decreasing the depths of those predicted to be over-estimated”; further demonstrated by how over/under-estimations are determined on pg. 5323, bullet points in the right column: “Under-estimate if the difference is more than t and the predicted depth value is less than the ground truth…Over-estimate if the difference is more than t and the predicted depth value is more than the ground truth”). It would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to have combined the over/under-estimation factors of Chawla with the system of Derbisz in order to improve the accuracy of the distance estimates of objects based on the actual distance determined by the system (Chawla, pg. 5321, 2nd to last paragraph: “method to correct the likely errors, showing that it can improve the depth map estimates”; pg. 5323, right column: “We formulated the error prediction as a per-pixel classification task, where the goal is to categorize the predicted depth estimations into correct, over-estimated, or under-estimated. To assign each pixel to a category, we check the absolute difference between the estimated depth D and the ground truth depth D*”).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure: Ding, D. S., & Chu, W. T. (2015, September). Weather-Adaptive Distance Metric for Landmark Image Classification. In Pacific Rim Conference on Multimedia (pp. 139-148). Cham: Springer International Publishing.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to EMMA E DRYDEN whose telephone number is (571)272-1179. The examiner can normally be reached M-F 9-5 EST.
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, ANDREW BEE can be reached at (571) 270-5183. 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.
/EMMA E DRYDEN/Examiner, Art Unit 2677 /ANDREW W BEE/Supervisory Patent Examiner, Art Unit 2677
Prosecution Insights