http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
- 中文 을 입력하시려면 zhongwen을 입력하시고 space를누르시면됩니다.
- 北京 을 입력하시려면 beijing을 입력하시고 space를 누르시면 됩니다.
RISS 인기검색어
- 검색키워드
- 저자 : Kirubakaran S. (검색결과 1 건)
- 무료
- 기관 내 무료
- 유료
-
Effective Anomaly Identification in Surveillance Videos Based on Adaptive Recurrent Neural Network
Arul U.,Arun V.,Rao T. Prabhakara,Baskaran R.,Kirubakaran S.,Hussan M. I. Thariq 대한전기학회 2024 Journal of Electrical Engineering & Technology Vol.19 No.3
Surveillance systems completed in true environment are of a solid nature. As the environment is uncertain and variable, care gradually becomes confusing when moving away from a stable and controlled environment. Evidence to distinguish stressful abnormalities in video surveillance is a problematic issue due to leakage, video screaming, contradictions and motives. Hence, in this paper, adaptive recurrent neural network is developed for anomaly detection from the videos. The projected technique is a combination of recurrent neural network and crystal structure algorithm. In the anomality detection, the video should be changed into frames. After that, the images should be enhanced for improving image quality. Once, the image quality is enhanced, the image background should be eliminated for achieving object detection. In the proposed technique, the region of interest is utilized to attain the object detection step in the images. The detected object images are used to tracking the object in the images by using the proposed classifer. To enhance the object tracking system, the feature extraction is a required topic in the system. Maximally stable extremal regions is used to extract the required features from the images. Finally, the proposed classifer is utilized to achieve anomaly detection based on object movement in the input images. The projected strategy is implemented and evaluated by performance metrices. It is contrasted with conventional techniques such as convolutional neural network-particle swarm optimization (CNN-PSO) and CNN respectively.
연관 검색어 추천
이 검색어로 많이 본 자료
활용도 높은 자료
