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and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114.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 submission filed on June 2, 2026 has been entered.
Response to Amendment
Receipt is acknowledged of claim amendments with associated arguments/remarks, received June 02, 2026. Claims 1-2, 4-11, 13-18, 21-24 are pending with amendments to Claims 1-2, 4, 8-11, 13, 17-18, 22. Claims 23-24 are new. Claims 3, 12, 19-20 were cancelled.
Response to Arguments
Applicant’s arguments, see Remarks, pg 9-10, filed 06/02/2026, with respect to the rejections of claims 1, 3, 6-9, 11-13, 16-17, 21-22 under 35 U.S.C. § 102 has been fully considered and, in light of the associated amendment, is persuasive. Therefore, the rejection has been withdrawn. However, upon further consideration, a new grounds of rejection is made as being unpatentable over Gomez et al (WO 2022/187341, disclosed by applicant in IDS 06/04/2025) in view of Fleisig et al (US 2021/0342585).
Applicant’s arguments, see Remarks, pg , filed 06/02/2026, with respect to the rejections of claims 2, 4-5, 10, 14, 15, 18 under 35 U.S.C. § 103 have been fully considered and are dependent upon the arguments posed under the prior art cited for the independent claims as discussed above. No additional arguments were presented specific to any of the cited art pertaining to the dependent claims.
All arguments were addressed.
Claim Rejections - 35 USC § 112(a) – New Matter
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claims 23, 24 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The following claims each contain subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for pre-AIA the inventor(s), at the time the application was filed, had possession of the claimed invention.
Claim 23 recites “wherein the at least one machine learning model comprises a recurrent neural network trained to detect, from the analyzed feature data, patterns of features associated with temporal anomalies for tracking at least one object from one or more objects.”
The applicant cites ¶¶ [0005]-[0010], [0046], [0051]-[0060] for support. The only recitation regarding an “anomaly” was identified in ¶ [0056], which states “The combination of the CNN 210 and the RNNs 212 can provide an improvement in image resolution and anomaly detection from the improved image resolution.” No further guidance was identified regarding what an “anomaly” is and interpreted based on its plain meaning. Further, the claim is directed to “temporal anomalies” and the limitation was not readily identified to be supported by the specification. It is unclear if a “temporal anomaly” would be with respect to changes in the frequency of obtaining data or changes occurring within the patterns of the objects over time from data that is received on a given temporal schedule. It is unclear as to the intent of the interpretation and the specification does not readily support “temporal anomalies” thereby resulting in a lack of written description and is therefore considered as a new matter.
Claim 24 recites “wherein the at least one machine learning model is trained to generate a classification label indicative of an object class and whether an object is stationary or moving, wherein the classification label is generated based on a comparison of pixel features at the two or more different time instances of the satellite input images, and wherein the prediction is generated based on one or more classification labels for the object generated by the at least one machine learning model.”
The applicant cites ¶¶ [0005]-[0010], [0046], [0051]-[0060] for support. However, support for the claim beginning with generating a label in the specification appears to be associated with the "health status of the hydrocarbon equipment (e.g., "defective", "healthy")” (specification ¶ [0006], [0075]) “or a label indicating severity of the defects (e.g., low, medium, high, severe) indicating operation of the equipment outside of a normal operational range” (specification ¶ [0043]). The specification was not readily identified to support the limitations claim of 24, thereby resulting in a lack of written description and is therefore considered a new matter.
Claim Rejections - 35 USC § 112(b)
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claim 8 is rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claim 8 has been amended to recite “wherein identifying the one or more objects in the environment comprises determining a difference in size and structures of pixels from the analyzed feature data for the satellite input images.”
The amendment creates uncertainty if “a difference in size and structures” is referring to the pixels (as claimed by amendment) or the objects represented by pixels (original interpretation based on specification, ¶ [0019]). Thus, Applicant has failed to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor regards as the invention. For purposes of examination, the claim is interpreted as the different in size and structures of objects represented by pixels.
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.
Claims 1, 6-9, 11, 13, 16-17, 21, 23 are rejected under 35 U.S.C. 103 as being unpatentable over Gomez et al (WO 2022/187341, disclosed by applicant in IDS 06/04/2025) in view of Fleisig et al (US 2021/0342585).
Regarding Claim 1, Gomez et al teach a method for analyzing and correcting satellite images representing an environment that includes hydrocarbon equipment (method 1600 for receiving spatial and temporal satellite image data of hydrocarbon production sites for analyzing gas flaring equipment; Fig 16 and ¶ [00203], [00205], [00212]), the method comprising:
receiving, by a computer system, satellite input images representing the environment captured by one or more satellites at different time instances (the computing system receives satellite data of a region of interest that includes multiple hydrocarbon production sites and includes data collected over time, block 1610; Fig 16 and ¶ [00203]-[00205], [00209]);
receiving, by a communication network coupled to the computer system, environmental data that represents a state of the environment (the computing system may receive weather data for the region of interest to analyze atmospheric conditions of the multiple hydrocarbon production sites, block 1630; Fig 16 and ¶ [00203]-[00204], [00206]);
determining at least one feature for extracting from the satellite input images (from the satellite data, one or more flares are identified from the one or more of the multiple hydrocarbon production sites, block 1640; Fig 16 and ¶ [00203]-[00205]);
applying, based on determining the at least one feature and the state of the environment determined from the environmental data (from the satellite data and at least a portion of the additional (weather) data, one or more flares are identified from the one or more of the multiple hydrocarbon production sites, block 1640; Fig 16 and ¶ [00203]-[00205]), one or more image processing functions to adjust pixels of the satellite input images for extraction of the at least one feature (the satellite and additional relevant (weather) image data is analyzed at the pixel-level in the machine learning model 1200 to detect thermal channel data, indicative of a pixel representing a fire and characterized to represent a lit flare, including corrections to the image data for more accurate analysis in the renderings; Fig 12 and ¶ [00141]-[00146], [00156]-[00157]);
analyzing, by at least one machine learning model to the compensated one or more pixels, feature data including the at least one feature that is extracted from two or more different time instances of the satellite input images (from the satellite data acquired over time (¶ [00205], [00209]), the trained machine learning model (Fig 12 and ¶ [00156]) identifies one or more flares at one or more of the multiple hydrocarbon production sites to determine if the gas flare is lit, block 1640; Fig 16 and ¶ [00207]-[00209]);
identifying, based on the analyzed feature data and by the at least one machine learning model, one or more objects from the hydrocarbon equipment in the environment, the identifying being based on the two or more different time instances of the satellite input images (from the analyzed data, the trained machine learning model (Fig 12 and ¶ [00156]) identifies one or more flares at one or more of the multiple hydrocarbon production sites as lit or unlit at each time point included in the analysis to identify change in lit status, block 1640; Fig 16 and ¶ [00205], [00207]-[00209]); and
generating, based on the analyzed feature data and the identified one or more objects, a prediction for an object among the one or more objects in the environment (from the satellite data acquired over time (¶ [00205], [00209]), the trained machine learning model classifies (predicts) the one or more flares at one or more of the multiple hydrocarbon production sites as intermittent or continuous lit, block 1640; Fig 16 and ¶ [00207]-[00209]).
Although Gomez et al teaches applying one or more image processing functions to the satellite input images (the satellite and additional relevant (weather) image data is analyzed at the pixel-level in the machine learning model 1200 to detect thermal channel data, indicative of a pixel representing a fire and characterized to represent a lit flare, including corrections to the image data for more accurate analysis in the renderings; Fig 12 and ¶ [00141]-[00146], [00156]-[00157]), Gomez et al does not explicitly teach compensating one or more pixels of the satellite input images to remove noise associated with in one or more of the satellite input images.
Fleisig et al is analogous art pertinent to the technological problem addressed in the current application and teaches compensating one or more pixels of the satellite input images to remove noise associated with in one or more of the satellite input images (a satellite 55 image is organized as a pixel data as a pixel map (¶ [0023], [0038]) and pixels undergo erosion filtering with a raster phase component 36 to remove noise from the satellite image, operation 104; Fig 1, 2, 4, 11 and ¶ [0088]-[0089], [0099], [0123]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the current application to combine the teachings of Gomez et al with Fleisig et al including compensating one or more pixels of the satellite input images to remove noise associated with in one or more of the satellite input images. By removing noise and artifacts at a pixel level, edge detection is performed straightforward and efficiently, thereby resulting in higher accuracy in the correct identification of objects, as recognized by Fleisig et al (¶ [0004], [0123]).
Regarding Claim 6, Gomez et al in view of Fleisig et al teach the method of claim 1 (as described above), wherein the object is a piece of the hydrocarbon equipment (Gomez et al, the gas flaring equipment of the one or more hydrocarbon production sites is identified; ¶ [00205], [00209]), and the prediction comprises a status indicator for the piece of the hydrocarbon equipment (Gomez et al, the trained machine learning model classifies (predicts) the one or more flares of the gas flaring equipment at one or more of the multiple hydrocarbon production sites as intermittent or continuous lit, block 1640; Fig 16 and ¶ [00207]-[00209]), the status indicator representing a health status of the piece of the hydrocarbon equipment (Gomez et al, the status (lit or unlit) of each of the one or more flares of the gas flaring equipment may include determining a reason for an unexpected unlit flare, such as reasons associated with weather, regulatory or operational status; ¶ [00215]).
Regarding Claim 7, Gomez et al in view of Fleisig et al teach the method of claim 6 (as described above), further comprising: providing the prediction comprising the status indicator for the piece of the hydrocarbon equipment in the environment to a system configured to monitor the environment (Gomez et al, the reason for a status (lit or unlit) of each of the one or more flares of the gas flaring equipment may be determined, including reasons associated with weather, regulatory or operational status; ¶ [00215]).
Regarding Claim 8, Gomez et al in view of Fleisig et al teach the method of claim 1 (as described above), wherein identifying the one or more objects in the environment comprises determining a difference in size and structures of pixels (interpreted as the size and structure of the object represented by pixels, not pixels of different sizes and structures, specification ¶ [0060]) from the analyzed feature data for the satellite input images (Gomez et al, the satellite image can be analyzed at each pixel by the trained ML model to determine characteristics of a plume (smoke related to a lit flare) and may be characterized based on size, shape (structure) of the object (plume); Fig 12 and ¶ [00162]-[00163]).
Regarding Claim 9, Gomez et al in view of Fleisig et al teach the method of claim 1 (as described above), wherein analyzing the feature data by the at least one machine learning model (Gomez et al, one or more machine learning model 1200 is used for analyzing the satellite image data; Fig 12 and ¶ [00156], [00167]) comprises:
generating, by a convolutional neural network (CNN) of the at least one machine learning model, subsets of the satellite input images (Gomez et al, a CNN may be utilized for pixel-wise segmentation; Fig 12 and ¶ [00157]), wherein a subset of the satellite input images share a plurality of topological features in the feature data determined by the CNN (Gomez et al, the pixel-wise classification (by the CNN) of the segmented regions is used to identify, map and classify different object types; Fig 12 and ¶ [00157]);
determining, by one or more recurrent neural networks (RNN) of the at least one machine learning model (Gomez et al, a long short-term memory (LSTM) RNN may be used as the ML model 1200; Fig 12 and ¶ [00167]-[00168]), one or more patterns over the different time instances from the feature data for the subsets of the satellite input images (Gomez et al, the satellite image data are input as time series data to the LSTM RNN to detect temporal patterns regarding the flare; Fig 12 and ¶ [00167]-[00168]);
identifying one or more objects in the environment based on the one or more patterns (Gomez et al, the patterns are used to recognize the flare as lit or unlit based on the detected temporal pattern by the LSTM RNN; Fig 12 and ¶ [00167]-[00168]); and
generating the prediction based on the one or more patterns for the one or more objects in the environment from the feature data (Gomez et al, the LSTM RNN to detect temporal patterns regarding the flare; Fig 12 and ¶ [00167]-[00168]).
Regarding Claim 11, Gomez et al in view of Fleisig et al teach the method of claim 1 (as described above), wherein the one or more image processing functions comprises at least one of (i) denoising, (ii) filtering, (iii) contrast adjustment, (iv) position alignment, (v) downsampling, (vi) up-sampling, or (vii) edge enhancement, of the pixels in the satellite input images (Gomez et al, false positives may be identified from intermittent flares based on flare satellite imagery and environment data, and the satellite images may be filtered based on atmospheric conditions indicated by the weather data; ¶ [00142], [00206]), and
wherein the environmental data (Gomez et al, environmental data from environmental satellites; ¶ [00198]) comprises at least one of (i) infrared data, (ii) simulation data, (iii) weather conditions, (iv) ground truth measurements, (v) historical data, (vi) geological data, or (vii) climate data, of the environment (Gomez et al, the computing system may receive weather data for the region of interest to analyze atmospheric conditions of the multiple hydrocarbon production sites, block 1630; Fig 16 and ¶ [00206], [00212]).
Regarding Claim 13, Gomez et al teach a system for analyzing and correcting satellite images representing an environment that includes hydrocarbon equipment (system 110 for well testing and equipment in environment 101 and/or marine environment 102; ¶ [0034]-[0036]), the system comprising: at least one processor (processor 112; Fig 1 and ¶ [0036]); and a memory storing instructions (memory 114 storing instructions 116; Fig 1 and ¶ [0036]) that, when executed by the at least one processor, cause the at least one processor to perform operations (processor 112 executes instructions 116 to examine operations in environment 101, 102; Fig 1 and ¶ [0036]) comprising: steps identical to claim 1 (as described above).
Regarding Claim 16, Gomez et al in view of Fleisig et al teach the system of claim 13 (as described above), wherein the limitations are claimed identical to claim 6 (as described above).
Regarding Claim 17, Gomez et al teach one or more non-transitory computer readable media storing instructions (memory 114 storing instructions 116; Fig 1 and ¶ [0036]) to analyze and correct satellite images representing an environment that includes hydrocarbon equipment (system 110 for well testing and equipment in environment 101 and/or marine environment 102; ¶ [0034]-[0036]), the instructions, when executed by at least one processor, configured to cause the at least one processor to perform operations (processor 112 executes instructions 116 to examine operations in environment 101, 102; Fig 1 and ¶ [0036]) comprising: steps identical to claim 1 (as described above).
Regarding Claim 21, Gomez et al in view of Fleisig et al teach the method of claim 1 (as described above), wherein the prediction for the object comprises a position in the environment at a time after the two or more different time instances (Gomez et al, spatial-temporal imaging assessed by the ML model may be N clips over time (thereby accounting for time and position at a third time used in the time series data; ¶ [00179], [0209]-[00214]).
Regarding Claim 23, Gomez et al in view of Fleisig et al teach the method of claim 1 (as described above), wherein the at least one machine learning model comprises a recurrent neural network trained to detect, from the analyzed feature data, patterns of features associated with temporal anomalies for tracking at least one object from one or more objects (Gomez et al, a long short-term memory (LSTM) RNN may be used as the ML model 1200 to detect temporal patterns regarding the flare recognize the flare as lit or unlit (anomaly), thereby tracking the temporal pattern of the flare; Fig 12 and ¶ [00167]-[00168]).
Claims 2, 14, 18 are rejected under 35 U.S.C. 103 as being unpatentable over Gomez et al (WO 2022/187341, disclosed by applicant in IDS 06/04/2025) in view of Fleisig et al (US 2021/0342585) and Avidan et al (US 2020/0342242, cited in Non-Final Rejection 12/19/2025).
Regarding Claim 2, Gomez et al in view of Fleisig et al teach the method of claim 1 (as described above), including the at least one machine learning model is trained to generate predictions of objects from the hydrocarbon equipment in the environment (Gomez et al, from the satellite data acquired over time (¶ [00205], [00209]), the trained machine learning model classifies (predicts) the one or more flares at one or more of the multiple hydrocarbon production sites as intermittent or continuous lit, block 1640; Fig 16 and ¶ [00207]-[00209]).
Gomez et al does not teach to generate predictions of objects from the hydrocarbon equipment in the environment, with the method comprising: generating, by a simulation configured to generate synthetic satellite images of the environment and using the environmental data, a training example comprising the generated synthetic satellite images; applying, by the at least one machine learning, the one or more image processing functions to the training example to generate a training set of adjusted images, wherein an adjusted image from the training set of adjusted images comprises pixels adjusted by the one or more image processing functions; comparing the training set of adjusted images from the generated synthetic satellite images of the environment to a set of adjusted images generated from the satellite input images of the environment; and updating, based on a comparison of the training set of adjusted images and the set of adjusted images, one or more parameters of the at least one machine learning model.
Avidan et al is analogous art pertinent to the technological problem addressed in this application and teaches generating, by a simulation configured to generate synthetic satellite images of the environment and using the environmental data (a synthetic data platform 107 is used to generate records for the geographic database 109, which includes synthetic image records 1009; Fig 1 and ¶ [0075], [0099]-[0100]), a training example comprising the generated synthetic satellite images (remote sensing satellite photography is used to generate records for the geographic database 109, which includes the generated synthetic image records 1009; Fig 1 and ¶ [0075], [0099]-[0100]);
applying, by the at least one machine learning model (the machine learning system 103; Fig 1 and ¶ [0075]), the one or more image processing functions to the training example to generate a training set of adjusted images (the machine learning system 103 processes an input image to process the synthetic image data generated to detect objects and labeling the images; Fig 1 and ¶ [0075]-[0077]),
wherein an adjusted image from the training set of adjusted images comprises pixels adjusted by the one or more image processing functions (the machine learning system 103 processes a portion of an input image in a given grid or receptive field and may perform edge detection of geographic features is detected (understood to be performed with pixel analysis) in the generation of synthetic image data; ¶ [0075]-[0076], [0086]);
comparing the training set of adjusted images from the generated synthetic satellite images of the environment to a set of adjusted images generated from the satellite input images of the environment (the mapped geographic features stored in the geographic database 109 are used to facilitate the generation of the synthetic image data used for machine learning; Fig 1 and ¶ [0076]); and
updating, based on a comparison of the training set of adjusted images and the set of adjusted images, one or more parameters of the at least one machine learning model (the machine learning model may be evaluated (updated) based on the labeled synthetic image data and labeled ground truth examples; Fig 1 and ¶ [0031]-[0032], [0037], [0081]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the current application to combine the teachings of Gomez et al in view of Fleisig et al with Avidan et al including generating, by a simulation configured to generate synthetic satellite images of the environment and using the environmental data, a training example comprising the generated synthetic satellite images; applying, by the at least one machine learning, the one or more image processing functions to the training example to generate a training set of adjusted images, wherein an adjusted image from the training set of adjusted images comprises pixels adjusted by the one or more image processing functions; comparing the training set of adjusted images from the generated synthetic satellite images of the environment to a set of adjusted images generated from the satellite input images of the environment; and updating, based on a comparison of the training set of adjusted images and the set of adjusted images, one or more parameters of the at least one machine learning model. By generating synthetic image data and using such data to evaluate, validate and retrain a machine learning model, an ample amount of training data may be generated and used by the model, thereby improving the generalizability and prediction accuracy for rare events but necessary for training the model for identifying such events, as recognized by Avidan et al (¶ [0032]).
Regarding Claim 14, Gomez et al in view of Fleisig et al teach the system of claim 13 (as described above), wherein the limitations are claimed identical to claim 2 (as described above).
Regarding Claim 18, Gomez et al in view of Fleisig et al teach the one or more non-transitory computer readable media of claim 17 (as described above), wherein the limitations are claimed identical to claim 2 (as described above).
Claims 4, 5, 15 are rejected under 35 U.S.C. 103 as being unpatentable over Gomez et al (WO 2022/187341, disclosed by applicant in IDS 06/04/2025) in view of Fleisig et al (US 2021/0342585) and Wen et al (Localization and Perception for Control and Decision Making of a Low Speed Autonomous Shuttle in a Campus Pilot Deployment).
Regarding Claim 4, Gomez et al in view of Fleisig et al teach the method of claim 1 (as described above), including the analyzed feature data from the satellite input images (Gomez et al, from the satellite data acquired over time (¶ [00205], [00209]), the trained machine learning model (Fig 12 and ¶ [00156]) identifies one or more flares at one or more of the multiple hydrocarbon production sites to determine if the gas flare is lit, block 1640; Fig 16 and ¶ [00207]-[00209]).
Gomez et al in view of Fleisig et al does not teach wherein the object is an autonomous vehicle and the prediction for the object comprises a trajectory comprising a plurality of predicted positions for the autonomous vehicle.
Wen et al is analogous art pertinent to the technological problem addressed in this application and teaches the object is an autonomous vehicle (an autonomous vehicle can be tracked in satellite images; Fig 1, 2, 8 and Real World Experiments ¶ 1, SLAM Evaluation ¶ 2) and the prediction for the object comprises a trajectory (a trajectory is predicted for the autonomous vehicle based on the SLAM analysis; Fig 8, 11, 12 and SLAM Evaluation ¶ 2-3, Real Time Path Following Performance) comprising a plurality of predicted positions for the autonomous vehicle (the trajectory includes the desired path trajectory and the expected predicted positions (which were then compared to actual path); Fig 8, 11-14 and SLAM Evaluation ¶ 2-3, Real Time Path Following Performance).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the current application to combine the teachings of Gomez et al in view of Fleisig et al with Wen et al including wherein the object is an autonomous vehicle and the prediction for the object comprises a trajectory comprising a plurality of predicted positions for the autonomous vehicle. By using localization and path planning for an autonomous vehicle, a trajectory can be easily identified and tracked, resulting in predictable expected paths of an autonomous vehicle and assist in tracking of multiple objects, as seen in aerial images, as recognized by Wen et al (Introduction ¶ 2-3).
Regarding Claim 5, Gomez et al in view of Fleisig et al and Wen et al teach the method of claim 4 (as described above), further comprising: providing the prediction comprising the trajectory (Wen et al, a trajectory is predicted for the autonomous vehicle based on the path planning; Fig 2, 3 and Path Tracking Model A. Path Generation, Error Calculation) to at least one of (i) a system configured to retrieve autonomous vehicles, or (ii) a computing device configured to monitor retrieval of autonomous vehicles, in the environment (Wen et al, the planned path trajectory is monitored to determine accuracy, as analyzed using a computing device; Fig 11-14 and SLAM Evaluation ¶ 2-3, Real Time Path Following Performance).
Regarding Claim 15, Gomez et al in view of Fleisig et al teach the system of claim 13 (as described above).
Gomez et al in view of Fleisig et al does not teach wherein the object is an autonomous vehicle and the operations further comprise: providing the prediction a trajectory to at least one of (i) a system configured to retrieve autonomous vehicles, or (ii) a computing device configured to monitor retrieval of autonomous vehicles, in the environment.
Wen et al is analogous art pertinent to the technological problem addressed in this application and teaches the object is an autonomous vehicle (an autonomous vehicle can be tracked in satellite images; Fig 1, 2, 8 and Real World Experiments ¶ 1, SLAM Evaluation ¶ 2) and the operations further comprise: providing the prediction a trajectory (a trajectory is predicted for the autonomous vehicle based on the SLAM analysis; Fig 8, 11, 12 and SLAM Evaluation ¶ 2-3, Real Time Path Following Performance) to at least one of (i) a system configured to retrieve autonomous vehicles, or (ii) a computing device configured to monitor retrieval of autonomous vehicles, in the environment (the planned path trajectory is monitored to determine accuracy, as analyzed using a computing device; Fig 11-14 and SLAM Evaluation ¶ 2-3, Real Time Path Following Performance).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the current application to combine the teachings of Gomez et al in view of Fleisig et al with Singh et al including wherein the object is an autonomous vehicle and the operations further comprise: providing the prediction a trajectory to at least one of (i) a system configured to retrieve autonomous vehicles, or (ii) a computing device configured to monitor retrieval of autonomous vehicles, in the environment. By using localization and path planning for an autonomous vehicle, a trajectory can be easily identified and tracked, resulting in predictable expected paths of an autonomous vehicle and assist in tracking of multiple objects, as seen in aerial images, as recognized by Wen et al (Introduction ¶ 2-3).
Claim 10 is rejected under 35 U.S.C. 103 as being unpatentable over Gomez et al (WO 2022/187341, disclosed by applicant in IDS 06/04/2025) in view of Fleisig et al (US 2021/0342585) and Huang et al (Meteorological Satellite Images Prediction Based on Deep Multi-scales Extrapolation Fusion, cited in Non-Final Rejection – 12/19/2025).
Regarding Claim 10, Gomez et al in view of Fleisig et al teach the method of claim 1 (as described above).
Gomez et al in view of Fleisig et al does not teach downsampling the satellite input images at a first resolution to a plurality of image datasets at a plurality of resolutions, wherein each resolution in the plurality of resolutions is lower than the first resolution, generating a first set of predictions for the plurality of image datasets; and combining, by a conditional generative adversarial network of the at least one machine learning model, the first set of predictions to a second set of predictions generated from the satellite input images at the first resolution, wherein the second set of predictions are generated at second resolution greater than the first resolution.
Huang et al is analogous art pertinent to the technological problem addressed in this application and teaches downsampling the satellite input images at a first resolution to a plurality of image datasets at a plurality of resolutions (the satellite images are scaled to several smaller sized by a down-sampling method; Fig 6-8 and 3.2 Method ¶ 1), wherein each resolution in the plurality of resolutions is lower than the first resolution, generating a first set of predictions for the plurality of image datasets (down-sampling of original satellite images results in several smaller scale sizes; Fig 6-8 and 3.2 Method ¶ 1); and
combining, by a conditional generative adversarial network of the at least one machine learning model (deep Multi-scales Extrapolation Fusion Generative Adversarial Network (GAN); Fig 6 and 3.2 Method ¶ 1), the first set of predictions to a second set of predictions generated from the satellite input images at the first resolution (the CGAN model is used to generate the prediction results of the satellite images by fusing the multi-scale prediction results; Fig 6, 9 and 3.2 Method ¶ 5), wherein the second set of predictions are generated at second resolution greater than the first resolution (the prediction images are generated at different scales and the scales are combined to generate the fused prediction satellite image; Fig 8, 9 and 3.2 Method ¶ 7).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the current application to combine the teachings of Gomez et al in view of Fleisig et al with Huang et al including downsampling the satellite input images at a first resolution to a plurality of image datasets at a plurality of resolutions, wherein each resolution in the plurality of resolutions is lower than the first resolution, generating a first set of predictions for the plurality of image datasets; and combining, by a conditional generative adversarial network of the at least one machine learning model, the first set of predictions to a second set of predictions generated from the input images at the first resolution, wherein the second set of predictions are generated at second resolution greater than the first resolution. By performing downsampling of the satellite images and using a spatiotemporal sequence for prediction of the different resolutions, realistic prediction images of the environment are gained from the satellite images, thereby allowing for spatiotemporal sequence predictions over a large area of environmental surface area, as recognized by Huang et al (1. Introduction ¶ 4).
Claim 22 is rejected under 35 U.S.C. 103 as being unpatentable over Gomez et al (WO 2022/187341, disclosed by applicant in IDS 06/04/2025) in view of Fleisig et al (US 2021/0342585) and Zhao et al (A Comprehensive Correction Method for Radiation Distortion of Multi-Strip Airborne Hyperspectral Images).
Regarding Claim 22, Gomez et al in view of Fleisig et al teach the method of claim 1 (as described above), including applying the one or more image processing functions comprises correcting reflectance values of the pixels of the satellite input images to reduce distortion in the satellite input images, the distortion associated with the state of the environment represented by the environmental data (Gomez et al, image processing of the satellite images can include filtering based on atmospheric conditions indicated by the weather data for the region of interest (reducing distortions causing false identifications) or correcting for spatial jittering caused by satellite geospatial positions (also to reduce distortions); ¶ [00142]-[00147], [00206]).
Gomez et al in view of Fleisig et al do not teach correcting reflectance values of the pixels of the input images.
Zhao et al is analogous art pertinent to the technological problem addressed in the current application and teaches correcting reflectance values of the pixels of the input images (airborne hyperspectral images are corrected for reflectance distortion; Fig 1, 2 and 3. Methodology ¶ 3, 3.2.3 Angle Normalization, 3.3 Consistency Adjustment between Multi-Strip Images ¶ 1-2).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the current application to combine the teachings of Gomez et al in view of Fleisig et al with Zhao et al including correcting reflectance values of the pixels of the input images. By correcting the aerial images for reflectance in a comprehensive manner, the geometry, terrain and elevation as well as environmental factors are considered and corrected, thereby improving the consistency for more accurate topological assessment, as recognized by Zhao et al (1. Introduction ¶ 6).
Claim 24 is rejected under 35 U.S.C. 103 as being unpatentable over Gomez et al (WO 2022/187341, disclosed by applicant in IDS 06/04/2025) in view of Fleisig et al (US 2021/0342585) and Matos-Carvalho et al (Static and Dynamic Algorithms for Terrain Classification in UAV Aerial Imagery).
Regarding Claim 24, Gomez et al in view of Fleisig et al teach the method of claim 1 (as described above), including the at least one machine learning model to analyze satellite input images (Gomez et al, the trained machine learning model classifies (predicts) the one or more flares of the gas flaring equipment at one or more of the multiple hydrocarbon production sites as intermittent or continuous lit from the satellite images, block 1640; Fig 16 and ¶ [00207]-[00209]),
Gomez et al in view of Fleisig et al do not teach wherein the at least one machine learning model is trained to generate a classification label indicative of an object class and whether an object is stationary or moving, wherein the classification label is generated based on a comparison of pixel features at the two or more different time instances of the satellite input images, and wherein the prediction is generated based on one or more classification labels for the object generated by the at least one machine learning model.
Matos-Carvalho et al is analogous art pertinent to the technological problem addressed in the current application and teaches at least one machine learning model is trained to generate a classification label indicative of an object class and whether an object is stationary or moving (machine learning model for terrain classification includes identification of static textures and dynamic textures, which are then classified (labeled), which is repeated to determine object dynamics (movement); Fig 5 and 2.3 Proposed System Model),
wherein the classification label is generated based on a comparison of pixel features at the two or more different time instances of the satellite input images (classification as static or dynamic includes multiple frames based on optical flow (temporal analysis) and performed with pixel analysis; Fig 5, 9 and 2.3 Proposed System Model, 2.5 Dynamic Textures), and
wherein the prediction is generated based on one or more classification labels for the object generated by the at least one machine learning model (object prediction based the classification incorporates the analysis of identifying whether the object was identified static or dynamic; Fig 5 and 2.3 Proposed System Model).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the current application to combine the teaching of Gomez et al in view of Fleisig et al with Matos-Carvalho et al including wherein the at least one machine learning model is trained to generate a classification label indicative of an object class and whether an object is stationary or moving, wherein the classification label is generated based on a comparison of pixel features at the two or more different time instances of the satellite input images, and wherein the prediction is generated based on one or more classification labels for the object generated by the at least one machine learning model. By analyzing aerial images for dynamic and static objects based on pixel analysis, better identification of objects, including borders, can be determined and by performing temporal analysis, classification be identified for dynamic objects, thereby improving the pattern recognition process for accurate object detection and classification, as recognized by Matos-Carvalho et al (1. Introduction ¶ 9).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Salman et al (US 2022/0262104, cited in Non-Final Rejection – 12/19/2025) teach a method and system for using a machine learning model to identify and classify objects from satellite images and label the probabilistically identified object concerning to equipment, operations and features, such as damage to the equipment based on temporal analysis.
Schmidt et al (US 2021/0398289, cited in Non-Final Rejection – 12/19/2025) teach a system and method that utilizes machine learning to analyze geospatial data for well pad detection and determining a probability of damage to the well pad from an environmental event.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to KATHLEEN M BROUGHTON whose telephone number is (571)270-7380. The examiner can normally be reached Monday-Friday 8:00-5:00.
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, John Villecco can be reached at (571) 272-7319. 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.
/KATHLEEN M BROUGHTON/Primary Examiner, Art Unit 2661
Prosecution Insights