Authors : Sanchit Sharma, Anshul Sharma and Nitish Ojha
Abstract: In this study, we are trying to recognize the irregular structure and vulnerable movement of people in the crowd to detect any anomaly in the situation by the movement of segmented particles. To accomplish this we are using a particle structure group in the image and observing its movement with the movement of the people. As the people move the particle density contract or expands according to the movement and speed of movement. The link between the contractions or expansion is mapped in the original image. When the recognized movement is too fast from the group of particles and if we can identify the person, we consider it a vulnerable object. These tests provide us with the modern analysis of too fast moving people in the crowd to recognize the hazardous situation. The examinations demonstrate that the proposed technique catches the progression of the group conduct effectively. In expansion, we have demonstrated that the social constrain approach beats comparative methodologies in light of immaculate optical stream.
Sanchit Sharma, Anshul Sharma and Nitish Ojha, 2017. Distance and Speed Based Anomaly Detection in Human Crowd Movement. Journal of Engineering and Applied Sciences, 12: 5467-5472.