Target Tracking Algorithm Based On Image Significance Detection Research

Due to the development of the economy, in the field of computational vision,moving target detection and target tracking is the core content of this field, mainlythrough behavior recognition. Image processing and other related knowledge toanalyze and understand the video stream. It is widely used in aerospace. Intelligentmonitoring and other fields has the very strong significance. In monitoring scenario,intelligent monitoring plays an important role in monitoring field, thro ugh theanalysis of intelligent video of abnormal behavior for the alarm and monitoring, theextremely important influence on social security. Target tracking in the intelligentmonitoring play an important role, by capturing the image target of interest, to track it,record the location of the object tracking and behavior and so on, and the records ofthe results were studied, have a far-reaching influence.At present, the domestic and foreign scholars put forward some of the moremature algorithm, there are mainly Point tracking, Nuclear tracking, Silhouettetracking. point tracking calculates the center of mass of tracking targets, by trackingpoint transformation between tracking, nuclear track through the target characteristicsof rectangular box to extract the core, and by calculating the displacement toimplement tracking, silhouette tracking target is often complicated geometry, throughthe segmentation of the target area to establish the target model, in each frame topredict to track the target method. Such as: to target profile modeling, but all of thetarget tracking algorithm are manually selected first frames to track the target, havecertain limitations.Significant regional detection technology is one of the highlights of the presentstudy, the technology has been widely used in image segmentation, imagecompression, content-based image retrieval, etc. Mainly based on the characteristicsof human visual attention to extract a scenario, the most important and the mostinteresting part of the largest amount of information, by the color of the image, prospects and the characteristics of regional segmentation, calculate the characteristic,at the same time also on the target detection and target tracking provide goodprospects of development and direction.This article uses the image significance detection target tracking, due to theimage detection algorithm is fast, is not sensitive to light, can be very good to extractthe image of significant areas, so by using the algorithm of image significance test ofthe first and second frame images of video streaming remarkable goal, and calculatethe image cancroids. Second, determine the second frame significantly targetcancroids position, with the first frame of the image target cancroids position notconsistent, then the target is locked for moving targets, or eliminate the targets. Dueto the first frame of the target tracking algorithm is manually select the target tracking,so when detect the target is a moving target, the first frame in the video to locate thetarget location, then tracking, can effectively avoid manually select target theamount of time, and be able to accurately find the target tracking.This article’s algorithm mainly through image significantly determine the motiontarget, for target tracking, it is suitable for different field, for example, video calls,intelligent transportation, At the same time, target tracking has a great role on theintelligent monitoring, through to the behavior of the target recognition, to monitor inthe scene pedestrians, Judgment in the scene pedestrian Is different from ordinarypeople, timely alarm processing, also can be used in smart phones, the target trackingalgorithm applied to the mobile phone to locate, ensure the personal safety, can alsobe applied in the field of intelligent transportation, through the analysis of vehicles,when an exception occurs, give an alarm in time.