# US 2025/0095374 A1 · Video analytics to detect animal abuse > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2025/0095374 A1 - Document ID: US20250095374A1 - Status: Published application - Kind: A1 - Application number: 18/641,963 - Filed: 2024-04-22 - Published: 2025-03-20 - Earliest claimed: 2020-02-28 - Assignees: ElectrifAi, LLC - Applicants: ElectrifAi, LLC - Inventors: Huiping Li; Michael Fox - Web page: https://patentgrep.com/patent/US20250095374A1 ## CPC - G06V 20/52: Surveillance or monitoring of activities, e.g. for recognising suspicious objects (recognising microscopic objects G06V20/69) - G06F 18/21: Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation - G06N 20/00: Machine learning - G06V 20/41: {Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items (segmenting video sequences G06V20/49)} - G06V 20/46: {Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames} - G06V 40/10: Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands - G06V 40/23: {Recognition of whole body movements, e.g. for sport training} - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract Video analytics are used to evaluate interactions between humans and animals and identify possible occurrences of animal abuse in an objective manner. The video analytics system processes successive video frames to identify objects of interest (e.g., humans, animals, tools/weapons, etc.), creates mathematical models of such objects (e.g., essentially stick-figure models), analyzes movements of such objects (e.g., the speed and/or directional motion of an object or portion of an object such as an arm or leg), determines mathematically and objectively whether or not the analyzed movements meet predetermined criteria for possible abuse (e.g., wherein the predetermined criteria can be mathematical models defining ethical and unethical movements), and outputs relevant information via a user interface (e.g., a list of possible abuse instances identifying the time and probability of possible abuse, from which the user can select an instance in order to view the corresponding video for human analysis). ## Claims ### Claim 1 A video analytics system for detecting animal abuse, the system comprising: a controller coupled to receive video data from at least one camera, the controller comprising at least one hardware processor coupled to at least one tangible, non-transitory computer readable medium containing instructions executable by the at least one processor to cause the controller to process successive video frames to identify object boundaries in each of the successive video frames, identify objects of interest including at least human body parts in the video frames based on the object boundaries including at least one of an arm, a leg, a torso, or a head, create mathematical models of such objects, analyze movements of such objects across the plurality of successive video frames to determine at least one of speed or directional motion of the body parts, determine mathematically and objectively whether or not the speed and/or directional motion of the body parts meet predetermined criteria for possible abuse of an animal, and, when an instance of possible animal abuse is detected, output information identifying the instance of possible animal abuse via a user interface. ### Claim 2 (depends on 1) A system according to claim 1, wherein the controller uses artificial intelligence or machine learning techniques in the analysis process, optionally wherein the controller receives feedback from a human regarding the identified instance of possible animal abuse and uses the feedback to enhance the artificial intelligence or machine learning techniques for future analysis. ### Claim 3 (depends on 1) A system according to claim 1, wherein the controller performs the video analytics substantially in real-time on each of a number of video streams or clips. ### Claim 4 (depends on 1) A system according to claim 1, wherein the controller identifies and models objects in the video via temporal and spatial modeling based on raw pixels. ### Claim 5 (depends on 4) A system according to claim 4, wherein the controller identifies boundaries in an image from the raw pixels and identifies features of objects based on the boundaries. ### Claim 6 (depends on 5) A system according to claim 5, wherein the controller tracks the features across multiple images to analyze movements of such objects and determines mathematically and objectively whether or not the analyzed movements meet predetermined criteria for possible abuse. ### Claim 7 (depends on 1) A system according to claim 1, wherein the objects further include at least one of an animal, a tool, or a weapon. ### Claim 8 (depends on 1) A system according to claim 1, wherein the mathematical models include stick-figure models including the body parts and wherein the controller analyzes movements of such objects by analyzing movements of the stick-figure models. ### Claim 9 (depends on 1) A system according to claim 1, wherein the predetermined criteria are mathematical models defining ethical and unethical movements. ### Claim 10 (depends on 1) A system according to claim 1, wherein the information identifying the instance of possible animal abuse includes at least one of a video clip showing the instance of possible animal abuse, a link to a portion of a video showing the instance of possible animal abuse, or information including a camera identifier, day, and time. ### Claim 11 (depends on 1) A system according to claim 1, wherein the controller assigns a probability score to each instance of possible animal abuse and allows the user to view video clips showing instances of possible animal abuse based on the probability scores. ### Claim 12 A method for detecting animal abuse in a video analytics system comprising a controller coupled to receive video data from at least one camera, the controller comprising at least one hardware processor coupled to at least one tangible, non-transitory computer readable medium containing instructions executable by the at least one processor to cause the controller to perform computer processes comprising: processing successive video frames to identify object boundaries in each of the successive video frames and to identify objects of interest including at least human body parts in the video frames based on the object boundaries including at least one of an arm, a leg, a torso, or a head; creating mathematical models of such objects; analyzing movements of such objects across the plurality of successive video frames to determine at least one of speed or directional motion of the body parts; determining mathematically and objectively whether or not the speed and/or directional motion of the body parts meet predetermined criteria for possible abuse of an animal; and when an instance of possible animal abuse is detected, outputting information identifying the instance of possible animal abuse via a user interface. ### Claim 13 (depends on 12) A system according to claim 12, wherein the controller uses artificial intelligence or machine learning techniques in the analysis process, optionally wherein the controller receives feedback from a human regarding the identified instance of possible animal abuse and uses the feedback to enhance the artificial intelligence or machine learning techniques for future analysis. ### Claim 14 (depends on 12) A system according to claim 12, wherein the controller performs the video analytics substantially in real-time on each of a number of video streams or clips. ### Claim 15 (depends on 12) A system according to claim 12, wherein the controller identifies and models objects in the video via temporal and spatial modeling based on raw pixels. ### Claim 16 (depends on 15) A system according to claim 15, wherein the controller identifies boundaries in an image from the raw pixels and identifies features of objects based on the boundaries. ### Claim 17 (depends on 16) A system according to claim 16, wherein the controller tracks the features across multiple images to analyze movements of such objects and determines mathematically and objectively whether or not the analyzed movements meet predetermined criteria for possible abuse. ### Claim 18 (depends on 12) A system according to claim 12, wherein the objects further include at least one of an animal, a tool, or a weapon. ### Claim 19 (depends on 12) A system according to claim 12, wherein the mathematical models include stick-figure models including the body parts and wherein the controller analyzes movements of such objects by analyzing movements of the stick-figure models. ### Claim 20 (depends on 12) A system according to claim 12, wherein the predetermined criteria are mathematical models defining ethical and unethical movements. ### Claim 21 (depends on 12) A system according to claim 12, wherein the information identifying the instance of possible animal abuse includes at least one of a video clip showing the instance of possible animal abuse, a link to a portion of a video showing the instance of possible animal abuse, or information including a camera identifier, day, and time. ### Claim 22 (depends on 12) A system according to claim 12, wherein the controller assigns a probability score to each instance of possible animal abuse and allows the user to view video clips showing instances of possible animal abuse based on the probability scores.