• Dimple Somaiya
• Sagar Virani
• Girish Mulchandani
• Viraj Daxini

The Internet of Things(IOT) devices like video cameras and sensors have small memories and less computational power. Traditional approaches for detecting, tracking and recognition of moving objects used only in the cloud computing. This approach suffered from high latency and more network bandwidth to transfer data into the cloud. We address this computing, IOT devices can’t handle high computation cost workload so we try to reduce computation cost of detection method which need to use CNN in video and simultaneously perform object detection/tracking. Multiple object detection, tracking and time series prediction are fundamental challenges in modern computer vision. Although deep learning has made significant strides in solving sum of the sub-problems, there are still many problems lacking satisfactory solutions, especially in real world application. Object detection and tracking are usually treated as two separate process.Which perform detection on every frame, so here I use tracking-by-detection pipeline by successfully detection on first frame and tracking associate by detection result. We will use well-known reference model of detection and tweaked the parameters extensively try to reduce load from cloud to edge which improve response time which indirectly reduce latency with better bandwidth availability, skipping detection from every frame and remove idle frame from computation which reduce computation cost with time budget. This output can be used for any further analytics like counting, direction finding, speed estimation etc. One of the fundamental application which is automotive use Edge-Computing as a single standalone processing power.

Edge Computing, Object Detection, tracking, YOLO

"OBJECT DETECTION, TRACKING AND PATTERN RECOGNITION MODEL USING EDGE COMPUTING AND MACHINE LEARNING ALGORITHM", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.9, Issue 4, page no.g117-g121, April-2022, Available :http://www.jetir.org/papers/JETIR2204618.pdf

"OBJECT DETECTION, TRACKING AND PATTERN RECOGNITION MODEL USING EDGE COMPUTING AND MACHINE LEARNING ALGORITHM", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.9, Issue 4, page no. ppg117-g121, April-2022, Available at : http://www.jetir.org/papers/JETIR2204618.pdf