Method and system for land encroachment detection and surveillance
원문보기
IPC분류정보
국가/구분
United States(US) Patent
등록
국제특허분류(IPC7판)
G06K-009/00
G06K-009/62
G08B-021/18
출원번호
16533033
(2019-08-06)
등록번호
11120259
(2021-09-14)
발명자
/ 주소
Hammad, Ali H
Arunachalam, Sekar
출원인 / 주소
Saudi Arabian Oil Company
대리인 / 주소
Leason Ellis LLP
인용정보
피인용 횟수 :
0인용 특허 :
0
초록▼
A system, a method, and a computer program for surveillance and area encroachment detection, including receiving satellite image data containing an image of a geospatial area, extracting features from the satellite image data of an object in the image, classifying the object in the image based on th
A system, a method, and a computer program for surveillance and area encroachment detection, including receiving satellite image data containing an image of a geospatial area, extracting features from the satellite image data of an object in the image, classifying the object in the image based on the extracted features, comparing the extracted features of the object to previously extracted features for the geospatial area, determining delta features between the extracted features of the object and the previously extracted features for the geospatial area, determining existence of an alarm event in the geospatial area based on the delta features, and transmitting an alarm event message to a communicating device.
대표청구항▼
1. A method for surveillance and area encroachment detection, the method comprising: receiving satellite image data containing an image of a geospatial area, the geospatial area comprising at least one critical zone and at least one permissive zone;extracting features from the satellite image data o
1. A method for surveillance and area encroachment detection, the method comprising: receiving satellite image data containing an image of a geospatial area, the geospatial area comprising at least one critical zone and at least one permissive zone;extracting features from the satellite image data of an object in the image;classifying the object in the image based on the extracted features;comparing the extracted features of the object to previously extracted features for the geospatial area;determining delta features between the extracted features of the object and the previously extracted features for the geospatial area;determining existence of an alarm event in the geospatial area based on the delta features, the existence of the alarm event being determined when at least a portion of the delta features is located in the at least one critical zone; andtransmitting an alarm event message to a communicating device when the existence of the alarm event is determined,wherein the alarm event determining determines a nonexistence of the alarm event when the delta features are located in the at least one permissive zone. 2. The method in claim 1, wherein the alarm event comprises an encroachment in the geospatial area. 3. The method in claim 1, wherein the alarm event comprises an encroachment in a zone in the geospatial area. 4. The method in claim 1, wherein the alarm event comprises adding or appearance of the object in the geospatial area. 5. The method in claim 1, wherein the alarm event comprises removal of the object from the geospatial area. 6. The method in claim 1, wherein the alarm event comprises a change to an attribute of the object. 7. The method in claim 1, further comprising receiving the previously extracted features for the geospatial area from a storage unit containing geographic information system data. 8. The method in claim 1, further comprising: receiving a training dataset for the geospatial area; andsupplying the training dataset to a deep learning neural network to train the deep learning neural network to determine the existence of the object in the geospatial area. 9. A non-transitory computer readable storage medium storing surveillance and area encroachment detection program instructions for causing an area encroachment detection and surveillance apparatus to analyze satellite image data and detect an alarm event, the program instructions comprising the steps of: receiving satellite image data containing an image of a geospatial area, the geospatial area comprising at least one critical zone and at least one permissive zone;extracting features from the satellite image data of an object in the image;classifying the object in the image based on the extracted features;comparing the extracted features of the object to previously extracted features for the geospatial area;determining delta features between the extracted features of the object and the previously extracted features for the geospatial area;determining existence of an alarm event in the geospatial area based on the delta features, the existence of the alarm event being determined when at least a portion of the delta features is located in the at least one critical zone; andtransmitting an alarm event message to a communicating device when the existence of the alarm event is determined,wherein the alarm event determining determines a nonexistence of the alarm event when the delta features are located in the at least one permissive zone. 10. The non-transitory computer readable storage medium in claim 9, the program instructions comprising the further step of: receiving a training dataset for the geospatial area; andsupplying the training dataset to a deep learning neural network to train the deep learning neural network to determine the existence of the alarm event in the geospatial area. 11. The non-transitory computer readable storage medium in claim 9, wherein the alarm event comprises an encroachment in the geospatial area. 12. The non-transitory computer readable storage medium in claim 9, wherein the alarm event comprises an introduction or appearance of an object in the geospatial area. 13. The non-transitory computer readable storage medium in claim 9, wherein the alarm event comprises change in the attributes of an object in the geospatial area. 14. The non-transitory computer readable storage medium in claim 9, wherein the alarm event comprises an encroachment in a zone in the geospatial area. 15. The non-transitory computer readable storage medium in claim 9, wherein the alarm event comprises removal of the object from the geospatial area. 16. The non-transitory computer readable storage medium in claim 9, the program instructions further comprising the step of receiving the previously extracted features for the geospatial area from a storage containing geographic information system data. 17. A surveillance and area encroachment detection system, the system comprising: an image data interface that receives, via a communication link to an image data source, satellite image data for an image of a geospatial area, the geospatial area comprising at least one critical zone and at least one permissive zone;an image processor having a deep neural network that extracts features from the satellite image data of an object in the image,classifies the object in the image based on the extracted features,compares the extracted features of the object to previously extracted features for the geospatial area,determines delta features between the extracted features of the object and the previously extracted features for the geospatial area, anddetects an alarm event in the geospatial area based on the delta features, the alarm event being detected when at least a portion of the delta features is located in the at least one critical zone; andarea encroachment detection and surveillance unit that transmits an alarm event message to a communicating device based on the detected alarm event,wherein the image processor determines a nonexistence of the alarm event when the delta features are located in the at least one permissive zone. 18. The system in claim 17, wherein the image processor comprises an object detector, a semantic segmentor, and an instance segmentor to compare extracted objects and their attributes at different times. 19. The system in claim 18, wherein the object detector predicts the object in the geospatial area. 20. The system in claim 18, wherein the object detector determines a bounding box and a probability score that the bounding box includes the object. 21. The system in claim 18, wherein the semantic segmentor associates each pixel in the satellite image data with a classification label. 22. The system in claim 18, wherein the instance segmentor comprises a Mask-RCNN. 23. The system in claim 18, wherein the instance segmentor marks out each object instance in the satellite image data.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.