A survey study on detecting and tracking objective methods

Abstract

The purpose of prosecution is segmentation feature a region of interest in a video scene and retention proposal, positioning and locking. Object detection and classification of objects are preceded steps for tracking an object in the image sequence. Object detection is done to check for objects in the video and find out exactly what the object. Then the detected object can be classified into different categories, such as people, vehicles, birds, clouds floating, swaying trees and other moving objects. Object Tracking is performed using spatial and temporal changes of items during a video monitoring, including presence, position, size, shape, etc. The object is used in several applications such as video surveillance, robot vision, monitoring traffic Video and Animation painting. This paper presents a brief overview of object detection, object classification and object tracking different algorithms available in the literature, including the analysis and comparative study of the various techniques used at various stages of prosecution.

Target Tracking based on KCF Combining with Spatio-Temporal Context Learning

“…The mean value of the position error is a commonly used evaluation criterion [11]. But this rough average evaluation method will bring assessment errors when the video sequence has a small amount of frame loss; they are not being able to fully reflect the real performance of the tracking method.…”

“…is the percentage of model in the t frame feature of target area to the study, and Equation (13) and Equation (14) can be used to calculate the target location of the t+1 frame in the local range of the target, which is shown in Equation (15). (16) where j denotes imaginary unit.…”

“…Most trackers are established on detection. The selection of training samples is a positive sample extraction at the centre of the target samples, whereas negative samples are based on the surrounding image [13]. Then, the positive sample is labeled 1, and the negative sample is labeled 0.…”

A robust tracking algorithm for a human‐following mobile robot

“…There are many different algorithms for object tracking. Object tracking can be divided into three categories, including point, kernel, and silhouette, as shown in Figure 1 [24]. In [26], Fang et al proposed a visual tracking method based on multi-cue information, they get a rough estimation of the target with optical flow, then adopted the part-based structure to optimize the result.…”

Real-time detection of moving objects in a video sequence by using data fusion algorithm

“…The optical flow method computes an optical flow image area based on the optical characteristics of a flow distribution image (Ramasubramanian et al, 2015). Sengar and Mukhopadhyay (2016) employed the normalized self-adaptive optical flow (NSOF) for efficient moving object area detection in a video sequence.…”