Detection and Tracking of a Moving Object Using Canny Edge and Optical Flow Techniques

Abstract

Aims: In the discipline of computer vision, detecting and tracking moving objects in a succession of video frames is a critical process. Image noise, complicated object motion and forms, and video real-time processing are some of the challenges faced by existing methods. Hence, they are computationally complex and susceptible to noise. This work utilized Canny Edge and Optical Flow (CE-OF) techniques for identifying and tracking moving objects in video files.
 Methodology: Video sequence datasets in Avi and Mp4 format from MathWorks and YouTube were used to evaluate the developed CE-OF technique. The video clip's frames were sampled several times and the frame rate display was calculated. The original images were converted to grayscale, preprocessed, and CE-OP was applied to identify and track the moving object. The results of the CE-OF and optical flow techniques in terms of accuracy, precision, false acceptance rate, false rejection rate, and processing time were obtained and compared. The performance of the developed technique was evaluated using accuracy, precision, false acceptance rate (FAR), false rejection rate (FRR) and processing time. The results obtained were 94.12%, 92.86%, 25.00%, 25.00%, and 19.51s for Mp4; and 93.33%, 90.91%, 20.00%, 20.00%, and 44.11s for Avi video 1 format, respectively.
 Conclusion: The developed CE-OF is a better competition in terms of accuracy and time compared with well-known techniques in the literature. The CE-OF technique performed better compared with the conventional methods in detecting and tracking a moving object. Therefore, it can be adopted in the designing of intelligent surveillance systems.

Temporal Action Segmentation in Sign Language System for Bahasa Indonesia (SIBI) Videos Using Optical Flow-Based Approach

“…Research on temporal segmentation is also closely related to motion detection or object movement tracking. The Canny Edge and Optical Flow (CE-OF) [22] method can detect and track moving objects with curation and precision above 90%. Apart from object tracking, Optical Flow can also be used to calculate movement speed [23], so it can potentially be used to extract information related to the speed of movement of signs, a nonmanual component in sign language.…”