Object Detection and Movement Tracking Using Tubelets and Faster RCNN Algorithm with Anchor Generation

Table 1

Notation with its description used in the proposed approach.

Symbol	Description
	Tracking confidence of object
	Detection score based on bounding box
	Bounding box
	Intersection over Union
	Bounding boxes detected
	Tubelet generated from video sequences
,	Visual features
	Relative movement of objects
and	Learning parameters
	Bounding box locations
,	Object movement
	Normalized movements of object
	Video sequence
	Object track
	Object detections
	Represents number of frames