Design and Implementation of Object Detection, Tracking, Counting and Classification Algorithms using Artificial Intelligence for Automated Video Surveillance Applications
share
Video Surveillance is very important and essential task for private and public organizations, sensitive areas to monitor safety, security issues and prevent crime. Especially in the field of surveillance system it has gained greater significance than ever before due to the recent terror activities taking place all over the world. Detection of objects precisely is vital for many applications such as person identification, abnormal activity detection, congestion analysis, military fields. Traditional surveillance was like capturing image or video from digital camera, store the data in a database, required information will be accessed manually from the database and it requires human operator to constantly monitor suspicious or threatening activities. Sometimes it may lead to loss of important information. In that particular cases automated video surveillance is very important. Current or automated surveillance system uses a digital technology to capture, store and process an image or video. Smart and intelligent surveillance is required to minimize the role of human operator and automatically detect, track, classify and monitor the suspicious or threatening activities. In this research work detection, tracking, counting and classification algorithms are implemented using matlab, DSP, FPGA and Artificial intelligence (AI) methods and described briefly. Feature of images is extracted using convolution neural network (CNN) using the concept of deep learning. YOLO based algorithm with GMM model by using the concepts of deep learning will give good accuracy for feature extraction and classification. AI combines SSD and Mobile Nets to perform efficient implementation of detection and tracking. This algorithm performs efficient object detection while not compromising on the performance. Further detection, tracking and classification algorithms are implemented for various applications under various constraints and environmental conditions.
READ FULL TEXT