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 .
Claims 1-20 are currently pending in U.S. Patent Application No. 18/917,476 and an Office action on the merits follows.
Priority
Applicant’s claim for the benefit of a prior-filed application under 35 U.S.C. 119(e) or under 35 U.S.C. 120, 121, 365(c), or 386(c) is acknowledged. Applicant has not complied with one or more conditions for receiving the benefit of an earlier filing date under 35 U.S.C. 119(e) (PRO 63/591,034 – “Connected Fly Light”) as follows:
The later-filed application must be an application for a patent for an invention which is also disclosed in the prior application. The disclosure of the invention in the parent/provisional application and in the later-filed application must be sufficient to comply with the requirements of 35 U.S.C. 112(a) or the first paragraph of pre-AIA 35 U.S.C. 112, except for the best mode requirement. See Transco Products, Inc. v. Performance Contracting, Inc., 38 F.3d 551, 32 USPQ2d 1077 (Fed. Cir. 1994).
The disclosure of the prior-filed provisional Application 63/591,034, fails to provide adequate support or enablement (meeting written description requirement) in the manner provided by 35 U.S.C. 112(a) or pre-AIA 35 U.S.C. 112, first paragraph for one or more claims of this application. More specifically, the provisional does not appear to disclose at least that comparing (Fig. 8 404 and 406) as recited in independent claim(s) 1/8/15. At best, the provisional disclosure in question considers historical data broadly ([0044]) and/or in [0049] “the system may implement a history based counting optimization, where a confidence score may be assigned to each detected fly based on the amount of time it has been consistently detected in history” (suggesting perhaps a confidence score based on track duration/FOV dwell time), but no disclosure of comparing one or more current detection boxes to one or more tracking boxes representing historical pest detections. Accordingly, all of pending claims 1-20 are understood to benefit from an Effective Filing Date no earlier than October 16, 2024.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claim(s) 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception, in particular an Abstract Idea falling under the (c) mental processes grouping (concepts performable in the human mind including an observation, evaluation, judgement, opinion), and/or alternatively the (a) mathematical concepts grouping (mathematical relationships, formulas or equations, and/or calculations (even if a series of calculations)), not ‘integrated into a practical application’ at Prong Two of Step 2A and without ‘significantly more’ at Step 2B.
Step 1: The claim(s) in question are directed to a computer implemented method for evaluating an image so as to determine and/or calculate an object/insect count. (Step 1: Yes).
Step 2A, Prong One: This part of the eligibility analysis evaluates whether the claim recites a judicial exception. As explained in MPEP 2106.04, subsection II, a claim “recites” a judicial exception when the judicial exception is “set forth” or “described” in the claim. Representative claim(s) 1/8/15 recite:
1) “determining output boxes based on comparing the one or more current detection boxes with historical detection boxes;”
2) “generating the output boxes; and”
3) “transmitting”/delivering/reporting “an object count based on the output boxes”.
Wherein 1-3 as broadly recited and accordingly permissibly interpreted (see MPEP 2173.01 and 2111.01), fall under the mental processes grouping (concepts performable in the human mind including an observation, evaluation, judgement, opinion), and/or alternatively the mathematical operations grouping as set forth in MPEP 2106.04(a)(2)(C):
A mathematical calculation is a mathematical operation (such as multiplication) or an act of calculating using mathematical methods to determine a variable or number, e.g., performing an arithmetic operation such as exponentiation. There is no particular word or set of words that indicates a claim recites a mathematical calculation. That is, a claim does not have to recite the word "calculating" in order to be considered a mathematical calculation. For example, a step of "determining" a variable or number using mathematical methods or "performing" a mathematical operation may also be considered mathematical calculations when the broadest reasonable interpretation of the claim in light of the specification encompasses a mathematical calculation.
Applicant may also consider Examples 47-49 of the 2024 Patent Eligibility Guidance (PEG): https://www.uspto.gov/sites/default/files/documents/2024-AI-SMEUpdateExamples47-49.pdf
The July 17, 2024 PEG identifies various process steps as drawn to the mathematical concepts Abstract Idea grouping – e.g. Example 47 claim 2 step(s) (b) (at page 7 describing the recited ‘discretizing’ as encompassing a mathematical concept e.g. rounding data values (that may also be performed mentally)) and (c) (interpreted so as to include mathematical calculations such as performing backpropagation and gradient descent algorithm(s)), in addition to Example 48 claim(s) 1 and 2 steps (b) (a ‘converting’ involving a mathematical operation using an STFT), (c) (determining (‘using’ a DNN) an ‘embedding’ on the basis of an explicitly recited formula), and (e) (‘applying binary masks’), and Example 48 claim 3 step(s) (c) (clustering using a k-means clustering algorithm) and (d) (binary masking clusters).
Examiner notes that while Applicant’s Specification at [0069] discloses an object detection (as pre-processing for obtaining a calculated and/or visually/mentally/manually determined count) to include “using a convolutional neural network or other suitable algorithm trained on pest image data”, such a use even if recited (which it is not) would at best be an additional element that does not preclude drawing any associated analysis/analyzing/detection under the mental processes grouping. Reference may be made to the 2024 PEG, Example 47 claim 2, wherein using an ANN did not preclude that anomaly detection and analysis of step(s) (d) and (e) from being drawn under the mental processes grouping at Prong One. See pages 6-7 of the above linked/referenced 2024 PEG document/Examples.
Dependent claims are similarly analyzed at least at Prong One since they inherit this/these same limitation(s) identified for the case of independent claim(s). Furthermore, claims such as e.g. claim(s) 4, additionally comprise limitations similarly/individually capable of being evaluated mentally/visually, and drawn under the mental processes grouping accordingly – i.e. even a comparison that involves “spatial overlap or appearance similarity” may still be performed visually/mentally. Bjerge et al. “An Automated Light Trap to Monitor Moths (Lepidoptera) Using Computer Vision-Based Tracking and Deep Learning”, relied upon in the prior-art based rejections below, further reinforces an understanding that associated imagery, and even bounding boxes themselves if it is persuasively asserted that they cannot be generated mentally/manually (e.g. manually generated training samples for a supervised learning training dataset may contain such boxes – particularly if the box itself is considered in class detection/IOU loss, e.g. YOLO), may be evaluated mentally/visually – page 12 “These videos were studied manually for each Track ID to evaluate the prediction against a visual species classification. Table 4 shows the algorithm’s estimate of the number of moth species and the established GT of the survey”. Regarding dependent claim 5, the claims do not recite what those first and second processing techniques constitute and/or how they necessarily differ, and permissible reading includes a difference with respect to human/evaluator attention/ consideration time (e.g. a user may choose to evaluate primarily those regions of the capture area (and associated sub-images) wherein the track/insect/object density is, was, or is expected to be, highest – i.e. closer to light source attracting the insects to be counted). (Step 2A, Prong One: Yes).
Step 2A, Prong Two: This part of the eligibility analysis evaluates whether the claim as a whole integrates the recited judicial exception into a practical application of the exception, distinct from the exception itself. This evaluation is performed by (1) identifying whether there are any ‘additional elements’ recited in the claim beyond the judicial exception, and (2) evaluating those ‘additional elements’ individually and in combination (weighed against the exception) to determine whether the claim as a whole integrates the exception into a practical application. See MPEP 2106.04(d). Examiner notes for consideration at Prong Two of 2A that MPEP 2106.05(a), (b), (c), and (e) generally concern elements that may be indicative of integration, whereas 2106.05(f), (g), and (h) generally concern elements that are not likely indicative of integration. As an additional note, ‘additional elements’ are generally limitations excluded from interpretation under the Abstract Idea groupings, and may comprise portions of limitations otherwise identified as falling under those Abstract Idea groupings of the 2019 PEG (e.g. any detection/determination/recognition that may be made mentally accompanied by the use of a neural network and/or generic computer hardware considered under the ‘apply it’ considerations of 2106.05(f)). Any ‘providing’/outputting broadly, and ‘collection’ of data (i.e. image acquisition(s)), be they images for training any learning model and/or data/images visually observable/ evaluated by a user/operator, also fail(s) to integrate at least in view of MPEP 2106.05(g) (extra-solution data gathering/output) and/or 2106.05(h) as ‘generally linking’ the exception to a field of use involving machine learning and/or imagery so acquired. Examiner also pre-emptively notes with respect to 2106.05(a), that ‘functioning of a computer’ (see fact pattern of Enfish, LLC v. Microsoft Corp., 822 F.3d 1327, 1336, 118 USPQ2d 1684, 1689 (Fed. Cir. 2016)) does not constitute operations that a general purpose computer may be programmed/configured to perform, since functioning of a computer instead concerns functions integral to the way computers operate (e.g. memory read-write for Enfish and virus scanning for Finjan). Regarding the claim(s) ‘as a whole’, the requirement for considering the claim as a whole stems from the fact that the judicial exception alone cannot provide the improvement, and any ‘additional elements’ are not evaluated in a vacuum separate from the weight of those directed to the exception (in further view of the Alice/Mayo’s roots in pre-emption). Consideration must be given to the degree/extent to which the apparent/disclosed improvement, as it is realized in recited claim language, is to the exception itself or otherwise distinct from it and captured by those limitations clearly serving as ‘additional elements’ after analysis at Prong One, in addition to how the ‘additional elements’ weigh in comparison to those limitations directed to the exception.
Reference may be made to the 08/04/2025 memo affirming analysis set forth in the 2024 PEG (https://www.uspto.gov/sites/default/files/documents/memo-101-20250804.pdf) and consistent with guidance to date. The most recent SME Memo(s) are available at: https://www.uspto.gov/patents/laws/examination-policy/subject-matter-eligibility and more specifically: https://www.uspto.gov/sites/default/files/documents/memo-desjardins.pdf
For the case of Desjardins, the claim(s) explicitly recited a limitation not drawn under/subsumed by the identified exception at Prong One, and realizing an improvement to the technical field of machine learning (serving for integration accordingly in view of 2106.05(a) – reciting an improvement to the way machine learning models are trained). The ARP’s decision in Desjardins also did not disturb the Board’s Prong One finding. The instant claims are unlike Desjardins however (do not concern any improvement to the technical field of training machine learning models), and read much more akin to an instance of ‘applying’ machine learning techniques (and/or mathematical operations e.g. a Hungarian Algorithm) to curate a set of bounding boxes, associated Track IDs, and ultimately derive/calculate a count based thereon. Even if counting numbers for various known/trained insect species is in itself useful/ practical – the utility of the exception itself does not serve for integration into a ‘practical application’ (see MPEP 2106.04(d)). Additional elements that include providing any notification and/or output of finally calculated statistics/counts, fail to serve for integration in view of MPEP 2106.05(g) and no additional elements outside of those directed to the exception itself, appear to explicitly/ specifically capture/recite any disclosed improvement in any technology and/or technical field (MPEP 2106.05(a)). With reference to MPEP 2106.05(a):
It is important to note, the judicial exception alone cannot provide the improvement. The improvement can be provided by one or more additional elements. See the discussion of Diamond v. Diehr, 450 U.S. 175, 187 and 191-92, 209 USPQ 1, 10 (1981))
Even when viewed in combination, the limited/minimal ‘additional elements’ present do not integrate the recited judicial exception into a practical application (Step 2A, Prong Two: No; Revised Step 2A: Yes [Wingdings font/0xE0] Step 2B).
Examiner requests Applicant’s assistance in providing a competing and compelling eligibility analysis, at Prong Two of Step 2A in particular, that explicitly identifies the improvement associated with Applicant’s invention (explicit or implied) (see MPEP 2106.05(a) sub-section II Improvements to any other technology or technical field – since the instant application does not concern “functioning of a computer” analogous to that of Enfish, (see above, and see also e.g. TJTM Technologies v Google, Appeal No. 2025-1218 (Fed. Cir. May 5, 2026) at page 5 citing Enfish)). The competing analysis should explicitly identify which limitations serve as the ‘additional elements’ realizing the improvement, and how these elements are not themselves subsumed under/within any exception. Applicant should also be advised, as Applicant’s representative(s) is/are likely aware, that the courts have declined to adopt the enumerated Abstract Idea groupings from 2019 Eligibility Guidance, and while the Examiner’s analysis does not rely on any Tentative Abstract Idea requiring approval (MPEP 2106.04(a)(3)), the instant claims are very arguably directed to a collection of data, analysis (even if one involving a curating/splitting/merging, etc., of tracks), and displaying/transmitting certain results (e.g. one or more counts) of the collection and analysis. Recently and in the realm of image analysis, Dental Monitoring SAS, v Align Technology, Inc., Appeal No. 2024-2270 (Fed. Cir. July 07, 2026), available at - https://www.cafc.uscourts.gov/opinions-orders/24-2270.OPINION.7-7-2026_2719362.pdf
Step 2B: This part of the eligibility analysis evaluates whether the claim as a whole amounts to ‘significantly more’ than the recited exception, i.e., whether any ‘additional element’, or combination of additional elements, adds an inventive concept to the claim. The considerations of Step 2A Prong 2 and Step 2B overlap, but differ in that 2B also requires considering whether the claims feature any “specific limitation(s) other than what is well-understood, routine, conventional activity in the field” (WURC) (MPEP 2106.05(d)). Such a limitation if specifically recited however, must still be excluded from interpretation under any of the Abstract Idea groupings. Step 2B further requires a re-evaluation of any additional elements drawn to extra-solution activity in Step 2A (e.g. gathering video/image(s)) – however no limitations appear directed to any novel collection per se. For at least the case of representative claim(s), both the obtaining/receiving and final providing/transmitting are generically recited, if not WURC. Applicant may consider Longitude Licensing Ltd. v. Google LLC, No. 24-1202, (Fed. Cir. April 30, 2025) (available at https://www.cafc.uscourts.gov/opinions-orders/24-1202.OPINION.4-30-2025_2506816.pdf) (see e.g. pages 7-9). While it is the MPEP that governs Examination and not necessarily case law (2019 marking a shift away from analysis attempting to identify analogous case law from a large and growing body of possibly pertinent case law examples), this opinion and those referenced therein (e.g. Recentive in particular) Recentive Analytics, Inc., v. Fox Corp., Appeal No. 2023-2437, 18 (Fed. Cir. Apr. 18, 2025) available at https://www.cafc.uscourts.gov/opinions-orders/23-2437.OPINION.4-18-2025_2500790.pdf serve to illustrate the manner in which claims that seek to apply broad classes of machine learning to a ‘new’ field of use, and/or claim limitations that do not explain/capture how a purported inventive concept/ improvement is actually achieved, are not likely to be determined eligible/enforceable. For clarity purposes, Examiner’s analysis does not rely upon any factual finding that any specifically recited (instead all are at a high level of generality) ‘additional elements’ constitute only that which is WURC (e.g. 2106.05(d) and Berkheimer memo of 2018) – but instead that the minimally present ‘additional elements’ fail to serve as ‘significantly more’ when considered at 2B, for the same reasons they fail at Prong Two of 2A – 2106.05(f) (implementation on generic computer hardware not constituting a particular machine as defined in 2106.05(b)), 2106.05(h) (linking to a field-of-use, i.e. wherein the images concern an insect trap/capture area, or other features that do not preclude those same images from being evaluated visually/mentally and/or utilized in calculations (even if computer assisted - MPEP 2106.04(a)(2) subsection C) based thereon), and/or 2106.05(g) for any outputting/delivery of calculated statistics if the claim language precludes doing so manually. Previously referenced Dental Monitoring SAS, v Align Technology, Inc., Appeal No. 2024-2270 (Fed. Cir. July 07, 2026) is of note, at page 12 – ““the relevant inquiry is not whether the claimed invention as a whole is unconventional or non-routine.” See BSG Tech LLC v. Buyseasons, Inc., 899 F.3d 1281, 1290 (Fed. Cir. 2018)”. Reference may also be made to the 2024 PEG describing that an improvement/ inventive concept (for ‘significantly more’ determination(s)) cannot be to the judicial exception itself (e.g. even a novel calculation/counting operation(s), as a calculation/counting per se, would be ineligible). (Step 2B: No).
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 of this title, 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.
1. Claims 1-4, 7-11 and 14-18 are rejected under 35 U.S.C. 103 as being unpatentable over Bjerge et al. “An Automated Light Trap to Monitor Moths (Lepidoptera) Using Computer Vision-Based Tracking and Deep Learning” (2021) and Chen et al. (US 2019/0130580 A1).
As to claim 1, Bjerge discloses a method (page 1 Abstract “the paper proposes detection and classification of species by recording images of live individuals attracted to a light trap. An Automated Moth Trap (AMT) with multiple light sources and a camera was designed to attract and monitor live insects during twilight and night hours. A computer vision algorithm referred to as Moth Classification and Counting (MCC), based on deep learning analysis of the captured images, tracked and counted the number of insects and identified moth species”) comprising:
receiving an image of a capture area (page 1 Abstract “Observations over 48 nights resulted in the capture of more than 250,000 images with an average of 5675 images per night”, page 3 section 1 “Our novel image processing pipeline incorporates the temporal dimension of image sequences”, Fig. 2 at page 4 of 18 “Figure 2. A picture of 3840 x 2160 pixels taken by the light trap of the light table and nine resting moths”, Fig. 3 Input Image, page 5 Section 2.2 “The MCC algorithm was composed by a number of sequential steps, where each image in the recording was analyzed as illustrated in Figure 3”, etc.,);
PNG
media_image1.png
370
652
media_image1.png
Greyscale
generating one or more current detection boxes associated with objects based on performing object detection on the image (Fig. 3 Analysis blob detection and bounding box assignment prior to/for subsequent tracking analysis, page 5 of 18 “The first step was to read an image from the trap segmented as black and white, followed by blob detection to mark a bounding box around each detected insect as described in Section 2.2.1. The position of each insect region in the image was estimated based on the center of the bounding box”, Fig. 4, Section 2.2.1 of pages 5-6, etc.,);
PNG
media_image2.png
290
906
media_image2.png
Greyscale
determining output boxes based on comparing the one or more current detection boxes with historical detection boxes (Fig. 3 Tracking prior to Classification, page 5 “The second step tracked multiple insects in the image sequence as described in Section 2.2.2. Tracking was important for recording the movement and behavior of the individual insect in the light trap and ensuring that it was only counted once during its stay in the camera’s field of view”, page 6 Section 2.2.2 “Tracking was used to reduce each visit of an insect in the camera field of view to one observation. However, we note that an individual insect could be counted again if it left the light trap and returned later during the night. … The position and size of each insect were estimated for every single frame, and tracking could therefore be solved by finding the optimal assignment of insects in two consecutive images. The Hungarian Algorithm [29] was our chosen method for finding the optimal assignment for a given cost matrix. In this application, the cost matrix should represent how likely it was that an insect in the previous image had moved to a given position in the current image. The cost function was defined as a weighted cost of distance and area of matching bounding boxes in the previous and current image … After a match with minimum cost, the entry in the current matrix was assigned a Track ID from the entry in the former. The found Track IDs and entries were stored and used in the upcoming iteration. … The insect assigned to a dummy could be used to determine which insect from the previous image had left, or which insect had entered into the current image”, see also page 11 Section 2.2.4 “An insect was only counted if it was observed in more than three consecutive images thus ignoring noise created by insects flying close to the camera”, etc.,);
generating the output boxes (finalizing at least a subset of those detected tracks/ associated BB’s in the aggregate, for input of those 128 x 128 crops into the CNN model, page 7 of 18 Section 2.2.3 “Based on the given camera setup the bounding boxes were finally resized approximately three times to a fixed window size of 128 x 128 x 3 as input for the customized CNN model”, etc.,; Examiner notes that no grounds of rejection under 112(b) is raised, because the Examiner understands the recited generating to reference Applicant’s Fig. 8 412 ([0074]), shaping permissible interpretation (MPEP 2173.01 and 2111.01 a plain meaning reading not inconsistent with Applicant’s Specification) a grouping/curation/filtering/finalization of output boxes (non-specific in terms of how those boxes must be transformed in such a generating), that individually exist (at least in some capacity) prior to that step of ‘generating’/curating a grouping/ finalized set of those – in other words, what is generated is a set of boxes for subsequent processing, based on existing boxes – which is why “the” output boxes (with basis previously established and referenced by the language ‘the’) can be ‘generated’ (an act otherwise associated with the beginning of existence)); and
transmitting an object count based on the output boxes (Fig. 3 Statistics block post classification (and ‘based on’ box and Track ID assignments of prior detection and tracking steps), and transmission for final output/results stage, page 5 Section 2.2. Counting and Classification of Moths, “The final step collected information and derived a summary of counted individuals of known moth species and unknown insects detected and tracked by the algorithm. The summary information was annotated to the image with highlighted marks of each insect tracks”, page 11 Section 2.2.4 “The detection of insects was summarized based on the number of counted insects, the number of moth species found, and the number of unknown insects found (i.e., unknown to the trained CNN algorithm). The statistics were updated as the images were analyzed in order to enable visual inspection of the result during processing”, etc.,).
PNG
media_image3.png
230
1302
media_image3.png
Greyscale
PNG
media_image4.png
96
980
media_image4.png
Greyscale
In the interest of compact prosecution, and in view of any argument that the language “generating the output boxes” in view of Applicant’s [0074] requires (arguably improper to read in – see MPEP 2111.01 subsection II.) other than minimally curating/transforming detected bounding boxes arriving at a set of so-called output boxes (with assigned Track IDs), for subsequent processing, and instead/additionally generating output boxes based on an addition/combination/aggregation of current detection boxes and/or additional boxes, Chen evidences the obvious nature of such a bounding box aggregation (Fig. 13 1325-1326, [0220] “For a current key frame, a final set of bounding boxes 1326 can be determined using detector bounding boxes (or "high confidence bounding boxes") produced by the deep learning system 1208 and foreground bounding boxes produced by the blob detection system 1204 for the current key frame. For example, the foreground bounding boxes 1224 generated by the blob detection system 1204 and the detector bounding boxes 1323 generated by the deep learning system 1208 (in some cases, filtered from the pre-processing module) are output to the bounding box aggregation engine 1325. The bounding box aggregation engine 1325 can aggregate the blob bounding boxes 1323 and the blob bounding boxes 1324 (which can include lists of bounding boxes BBDetector and BBBgSub, respectively) to produce the final set of bounding boxes 1326 for a current key frame. The final set of bounding boxes 1326 for a key frame can include a final list of bounding boxes, which can be denoted as BBFinal. In some examples, a status can also be determined for each of the bounding boxes in the final set of bounding boxes 1326. Each of the bounding boxes in the final set 1326 can represent a blob detected for the video frame”, [0221] “A first bounding box from the blob bounding boxes 1324 and a second bounding box from the detector bounding boxes 1323 can be determined to be associated with one another when a large portion of the area of the first or second bounding box belongs to an overlapping area of the first and second bounding boxes. The overlapping area can include an intersecting region, which includes a region that includes the overlapping portion of the first bounding box and the second bounding box”, [0228], etc.,) prior to and for the purposes of determining an object track (Fig. 13 1206). Chen further evidences the obvious nature of determining a final set of bounding boxes (using ‘different’ and possibly scenario specific ‘aggregation techniques’ [0228], etc.,), based on the bounding boxes associated with two input frames, in view of the manner(s) in which each set of bounding boxes 1323 (detector) and 1324 (blob) may comprise corroborating information serving to exclude/minimize otherwise false positive detections (see various ‘false positive’ disclosure of Chen, e.g. [0229], [0231], [0234], etc.,) (see also Bjerge at page 7 – also concerned with measuring, and reducing, False Positive detections defined as a same individual miscounted/counted multiple times/assigned a new/erroneous Track ID).
It would have been obvious to a person of ordinary skill in the art, before the effective filing date, to modify the system and method of Bjerge to further comprise a bounding box generation/aggregation aimed at reducing/minimizing false positive detections/Track ID assignments, based on one or more of those aggregation techniques, confidence indicators, overlap metrics, etc., of Chen, as taught/suggested therein, the motivation as similarly taught/suggested therein and readily apparent to PHOSITA broadly that such an aggregation may ensure any subsequently/finally calculated statistics based thereon are characterized by a sufficient/desired level of accuracy.
As to claim 2, Bjerge in view of Chen teaches/suggests the method of claim 1.
Bjerge in view of Chen further teaches/suggests the method wherein the output boxes are equivalent to an amount of one or more current detection boxes based on the amount of the one or more current detection boxes being greater than an amount of historical tracking boxes (Bjerge Track ID assignment identified above, in further view of that proposed modification in view of the teachings of Chen, wherein, the recited condition would necessarily occur as a consequence of at least two frames considered wherein no existing/active track is dropped/lost and at least one new track is assigned, or, the number lost/dropped/exiting FOV is lower than the number entering/newly assigned (either scenario would result in the condition recited) – see also e.g. Chen [0184] “For instance, one active tracker can only be mapped to one blob. All the other blobs (the blobs remaining from the multiple blobs that are not mapped to the tracker) cannot be mapped to any existing trackers. In such examples, new trackers will be created for the other blobs, and these new trackers are assigned the state "split-new."”).
As to claim 3, Bjerge in view of Chen teaches/suggests the method of claim 1.
Bjerge in view of Chen further teaches/suggests the method wherein the output boxes are equal to an amount of one or more current detection boxes and (Examiner interprets this ‘and’ to be a summation (output = current + additional), and not, output = current = additional) an amount of additional boxes based on the amount of the one or more current detection boxes being less than an amount of historical tracking boxes (Bjerge Track ID assignment identified above, in further view of that proposed modification in view of the teachings of Chen, wherein, the recited condition would necessarily occur as a consequence of all assigned Track IDs being supplied for subsequent processing, even for the instance that one or more Track IDs is not present in a current, but was present in a past frame, in other words an instance where an object’s/moth’s ‘presence of the observation’ does not include the current frame (or has been eliminated/culled/removed from the current frame), but that Track ID is nonetheless counted – which is disclosed by Bjerge as understood by the Examiner particularly in view of the manner in which Bjerge states at page 6 section 2.2.2 “Tracking was used to reduce each visit of an insect in the camera field of view to one observation. However, we note that an individual insect could be counted again if it left the light trap and returned later during the night”; see also Chen at [0218] “1208 can include a pre-processing module (not shown) that can filter out (remove) the bounding boxes from the list of detector bounding boxes (BBDetector) 1323 if the bounding boxes are associated with an NOI object type. In some examples, only a COI object type can be defined for objects that are of interest to the video analytics system 1200, and any detected object that is not part of the COI object type can be filtered out from the detector bounding boxes 1323” in further view of Bjerge at 2.2.4 identifying at least some detection instances that are ideally treated as noisy detections (warranting removal/elimination) – e.g. insects flying close to the camera but not present for three consecutive frames).
As to claim 4, Bjerge in view of Chen teaches/suggests the method of claim 1.
Bjerge discloses the method wherein comparing the one or more current detection boxes with historical detection boxes comprises matching current detection boxes with historical detection boxes based on spatial overlap (Bjerge Hungarian Algorithm and cost matrix based optimal assignment – in view of that Areacost of equation(s) (3) and (4)).
Chen further evidences the obvious nature of a bounding box comparison based on spatial overlap or appearance similarity ([0221] “The overlapping area can include an intersecting region, which includes a region that includes the overlapping portion of the first bounding box and the second bounding box”, [0226-0228] etc., concerning explicitly disclosed overlap, and Chen further suggests an appearance similarity should either of at least object size/dimensions, and/or object class, be considered a characteristic of ‘appearance’). Chen further discloses in e.g. [0228] that a threshold for percentage of overlap may be scaled based on situation specific confidences, expected object sizes, etc..
It would have been obvious to a person of ordinary skill in the art, before the effective filing date, to further modify the system and method of Bjerge in view of Chen, such that the bounding box comparison thereof includes, in conjunction with and/or as an alternative to, that comparison of 2.2.2, a comparison on the basis of alternative equivalents of spatial overlap and/or appearance similarity broadly, as taught/suggested by Chen, the motivation as similarly taught/suggested therein and readily apparent to PHOSITA broadly that such characteristics for determining a match/similarity, as known/established/predictable characteristics useable in a similarity determination characterized by a reasonable expectation of success (as evidenced by at least Chen), are no more than those Obvious to Try given the level of skill in the art (see MPEP 2143 Rationale (E) in view of 2141 and Basic/Graham Factual Inquiry (C)).
As to claim 7, Bjerge in view of Chen teaches/suggests the method of claim 1.
Bjerge in view of Chen teaches/suggests the method further comprising displaying a visualization of object count trends over time based on the output boxes associated with multiple images (Bjerge Table 4, 6, Figs. 8-9, etc., in view of section(s) 3.2 and 3.3 in particular “The seasonal dynamics of each of the eight species of moths detected by the MCC algorithm showed clear similarities with peak abundance during the last days of August 2019 (Figure 8). However, Autographa gamma observations were restricted to mostly a single night, while other species were frequent during several consecutive nights. Some species including the Hoplodrina complex exhibited a second smaller peak in abundance around 10 September 2019. This variation was mirrored to some extent in the weather patterns from the same period (Figure 9). The statistical models of the relationships with weather patterns revealed that abundance of all species was positively related to night temperatures, while three species also responded to wind speed, and another three species responded to both air humidity and wind speed (Table 6)”).
As to claim 8, this claim is the device claim corresponding to the method of claim 1 and is rejected accordingly. Regarding generic computer hardware (i.e. processor and memory – see Bjerge at page 3 section 2.1 Hardware Solution – Raspberry Pi 4 computer).
As to claims 9-11 and 14, these claims are the device claims corresponding to method claims 2-4 and 7 respectively, and are rejected accordingly.
As to claim 15, this claim is the non-transitory CRM claim corresponding to the method of claim 1 and is rejected accordingly.
As to claims 16-18, these claims are the non-transitory CRM claims corresponding to method claims 2-4 respectively, and are rejected accordingly.
2. Claims 5-6, 12-13 and 19-20 are rejected under 35 U.S.C. 103 as being unpatentable over Bjerge et al. “An Automated Light Trap to Monitor Moths (Lepidoptera) Using Computer Vision-Based Tracking and Deep Learning” (2021), in view of Chen et al. (US 2019/0130580 A1) and Rummelhard et al. “Conditional Monte Carlo Dense Occupancy Tracker”.
As to claim 5, Bjerge in view of Chen teaches/suggests the method of claim 1.
Bjerge fails to explicitly disclose the method further comprising based on a historical object distribution pattern derived from the historical detection boxes:
using a first processing technique for a first region of the capture area; and
using a second processing technique for a second region of the capture area.
Chen at the minimum suggests using different processing techniques 1208 vs 1204 (the claim does not recite what the processing techniques do, or how) for regions of a capture area e.g. Fig. 5, even if it is argued that both techniques are applied to same regions (also using the ‘first’ technique on the ‘second’ region is not precluded by the recited language, so long as the first technique is used on a so-called first region), and further discloses various embodiments regarding historical information that appear equivalent to a historical object distribution pattern broadly (e.g. a track comprising historical locations) ([0139] “The prediction of the location of the blob tracker in the current frame can be based on the location of the blob in the previous frame. A history or motion model can be maintained for a blob tracker, including a history of various states, a history of the velocity, and a history of location, of continuous frames, for the blob tracker, as described in more detail below”, [0147], [0165] “In some cases, a tracker is assigned with a unique ID, and a history of bounding boxes is kept. … When blobs (making up at least portions of objects) are detected from an input video frame, blob trackers from the previous video frame need to be associated to the blobs in the input video frame according to a cost calculation. …”, [0166], etc.,).
Furthermore, Rummelhard evidences the manner in which PHOSITA, in the context of determining an object/target track prediction, may consider the problem from an environment/ capture-area perspective, and less and/or in conjunction with (complementary to – hybrid approach) an object-level context (Abs, “The most common approaches involve the modeling of moving objects, through Detection And Tracking of Moving Objects (DATMO) methods. An alternative to a classic object model framework is the occupancy grid filtering domain. Instead of segmenting the scene into objects and track them, the environment is represented as a regular grid of occupancy, in which spatial occupancy is tracked at a sub-object level. In this paper, we present the Conditional Monte Carlo Dense Occupancy Tracker, a generic spatial occupancy tracker, which infers dynamics of the scene through an hybrid representation of the environment, consisting of static occupancy, dynamic occupancy, empty spaces and unknown areas”). Rummelhard further evidences the obvious nature of applying first and second (different - adaptive) processing techniques to different regions (grids/cells) in consideration of an occupancy distribution (Section 1 “A more adaptive method can still drastically decrease the motion representation dimension. As most cells, like empty or static ones, do not necessitate elaborated representation, important processing resources can be saved by taking this information into consideration. In the process presented by Danescu et al. [11], the idea to use a variable number of samples per cell, according to the occupancy estimated”, section II “The occupancy distribution can then be inferred from those hidden states. Besides presenting a neater distinction between static and dynamic parts, the main interest of this modification is to introduce a specific processing of dataless areas, excluding them from the velocity estimation and disabling their temporal persistence”, etc.,).
It would have been obvious to a person of ordinary skill in the art, before the effective filing date, to further modify the system and method of Bjerge in view of Chen, so as to apply first and second (different - adaptive) processing techniques to different regions (grids/cells) in consideration of an occupancy distribution as taught/suggested by Rummelhard, the motivation as similarly taught/suggested therein that the same would make use of processing resources more efficiently, by at least allocating less computational resources in processing those areas/regions known/recognized as empty spaces and/or historically less frequented areas of the capture(d) space.
As to claim 6, Bjerge in view of Chen and Rummelhard teaches/suggests the method of claim 5.
Bjerge in view of Chen and Rummelhard further teaches/suggests the method wherein the first processing technique or the second processing technique comprises threshold-based segmentation of cells associated with a contrast, edge-based detection, or color-based segmentation of cells (see Bjerge Otsu/segmentation disclosure in e.g. pages 5-6, Section 2.2.1 and with reference to Fig. 4).
As to claims 12-13, these claims are the device claims corresponding to method claims 5-6 respectively, and are rejected accordingly.
As to claims 19-20, these claims are the non-transitory CRM claims corresponding to method claims 5-6 respectively, and are rejected accordingly.
Additional References
Prior art made of record and not relied upon that is considered pertinent to applicant's disclosure:
Additionally cited references (see attached PTO-892) otherwise not relied upon above have been made of record in view of the manner in which they evidence the general state of the art.
Sittinger et al. ‘Insect Detect: An open-source DIY camera trap for automated insect monitoring’ (attached PTO-892 page 2 NPL Citation No. U) is of note, see e.g. page 6 disclosing “The object tracker is based on the Kalman Filter and Hungarian algorithm, which can keep track of a moving object by comparing the bounding box coordinates of the current frame with the object’s trajectory on previous frames. By assigning a unique ID to each insect landing on or flying above the flower platform, multiple counting of the same individual can be avoided as long as it is present in the frame”.
Patch et al. (US 2022/0361471 A1) discloses at e.g. [0005] “generating an N-frame history buffer for each of the one or more bounding boxes in the image; determining validity of each of the one or more bounding boxes in the image based on the N-frame history buffer of a respective bounding box of the one or more bounding boxes;”
Inquiry
Any inquiry concerning this communication or earlier communications from the examiner should be directed to IAN L LEMIEUX whose telephone number is (571)270-5796. The examiner can normally be reached Mon - Fri 9:00 - 6:00 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, Chan Park can be reached on 571-272-7409. 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.
/IAN L LEMIEUX/Primary Examiner, Art Unit 2669
Prosecution Insights