Action recognition in a video sequence
US-2018137362-A1 · May 17, 2018 · US
US11270562B2 · US · B2
| Field | Value |
|---|---|
| Publication number | US-11270562-B2 |
| Application number | US-202017020893-A |
| Country | US |
| Kind code | B2 |
| Filing date | Sep 15, 2020 |
| Priority date | Feb 18, 2020 |
| Publication date | Mar 8, 2022 |
| Grant date | Mar 8, 2022 |
A practical reading order for non-experts. Skip the full description unless you need deep technical detail.
What the patent document calls the invention.
A short plain-language summary of the technical disclosure.
Who owns or filed the patent and who is credited as inventor.
Filing, priority, publication, and grant dates set the timeline.
The legal scope of protection — read this for what is actually claimed.
Technology tags used to group this patent with similar filings.
Prior art links and similar publications in this corpus.
Official abstract text for this publication.
An efficient alert process is realized and effectiveness of surveillance is improved when surveilling various incidents that occur within a surveillance area. In a video surveillance system for detecting a plurality of incidents in a surveillance area using a video that captured the surveillance area, provided are a calculation unit which detects an image of a person and/or an object from the video, and calculates an incident feature quantity as a feature quantity representing a status of a detected person and/or a feature quantity of a detected object, and a determination unit which detects an occurrence of the incident based on the incident feature quantity, and determines subjects to be alerted of an occurred incident, which is a detected incident, from a type of the occurred incident and a determination accuracy of the occurred incident.
Opening claim text (preview).
The invention claimed is: 1. A video surveillance system for detecting a plurality of incidents in a surveillance area using a video that captured the surveillance area, comprising: a memory storing at least one program; and a processor, which when executing the at least one program, configures the processor to: detect an image of a person and/or an object from the video, and calculate an incident feature quantity as a feature quantity representing a status of a detected person and/or a feature quantity of a detected object; detect an occurrence of an incident based on the incident feature quantity, and determine subjects to be alerted of an occurred incident, which is a detected incident, from a type of the occurred incident and a determination accuracy of the occurred incident; and identify an event indicating a combination of the subjects to be alerted based on the type of the occurred incident and the determination accuracy of the occurred incident, and output the subjects to be alerted indicated in the identified event, wherein the subjects to be alerted include a responder to respond to the occurred incident in the surveillance area, and a manager to manage a status of the surveillance area, and a part of the event indicates a combination which excludes the manager from the subjects to be alerted. 2. The video surveillance system according to claim 1 , wherein, when an input to the video surveillance system is a moving image and human skeletal frame detection is to be used for determining an incident of a person, the processor is further configured to calculate a feature quantity representing a posture of a person calculated from an estimation result of a skeletal frame based on the human skeletal frame detection, and a feature quantity representing an amount of movement of a skeletal frame by concurrently using a previous image frame and a subsequent image frame, and using the feature quantities integrally. 3. The video surveillance system according to claim 1 , wherein the video surveillance system includes a storage unit which stores surveillance standard information to be referenced for outputting the subjects to be alerted based on the occurred incident and the determination accuracy of the occurred incident, and wherein the processor is further configured to determine surveillance importance prescribed for each type of the incident based on a surveillance importance determination standard set in the surveillance standard information. 4. The video surveillance system according to claim 3 , wherein the processor is further configured to determine an event from the occurred incident type or the surveillance importance, and the determination accuracy of the occurred incident, based on an event determination standard set to the surveillance standard information. 5. The video surveillance system according to claim 4 , wherein the processor is further configured to determine the subjects to be alerted from the event based on a subject to be alerted determination standard set to the surveillance standard information. 6. The video surveillance system according to claim 1 , wherein the video surveillance system detects an incident based on an interaction, which is a coordinated action between multiple persons in the surveillance area, by using a video that captured the surveillance area. 7. The video surveillance system according to claim 6 , wherein the processor is further configured to use one type or multiple types of distances between body parts between persons for detection of interaction regarding the feature quantity to be calculated based on a skeletal frame of the detected person. 8. The video surveillance system according to claim 6 , wherein the processor is further configured to estimate, from a detection result of a person and an article, which person is in possession of the article for detection of interaction, and determine an interaction between persons via the article based on that information. 9. A video surveillance method for detecting a plurality of incidents in a surveillance area using a video that captured the surveillance area, comprising the steps of: detecting an image of a person and/or an object from the video, and calculating an incident feature quantity as a feature quantity representing a status of a detected person and/or a feature quantity of a detected object; detecting an occurrence of the incident based on the incident feature quantity; determining subjects to be alerted of an occurred incident, which is a detected incident, from a type of the occurred incident and a determination accuracy of the occurred incident; and identifying an event indicating a combination of the subjects to be alerted based on the type of the occurred incident and the determination accuracy of the occurred incident, and outputting the subjects to be alerted indicated in the identified event, wherein the subjects to be alerted include a responder to respond to the occurred incident in the surveillance area, and a manager to manage a status of the surveillance area, and a part of the event indicates a combination which excludes the manager from the subjects to be alerted.
Alarm destination chosen according to type of event, e.g. in case of fire phone the fire service, in case of medical emergency phone the ambulance · CPC title
Event detection · CPC title
using feature-based methods · CPC title
Recognition of a predetermined image pattern or behaviour pattern indicating theft or intrusion · CPC title
Recognition of whole body movements, e.g. for sport training · CPC title
Related publications grouped by family.
Answers are generated from the same data shown on this page.