Font Size: a A A

Based On The Background Of The Shadow Vehicle Road Features Eliminate Algorithms

Posted on:2013-11-29Degree:MasterType:Thesis
Country:ChinaCandidate:X Q KongFull Text:PDF
GTID:2248330392953311Subject:Software engineering
Abstract/Summary:PDF Full Text Request
As China’s vehicle market expands in a large scale, and the conflicts betweentraffic jam, vehicles and drivers stand out, I believe smart traffic in cities in China,especially medium and large ones, is a promising research and a practical project.The author has made substantial research in the national and international smarttraffic system. The most important part of a smart traffic system is the identificationof vehicles. The development and the application of other systems all rely on theidentification of vehicles and license plate. In order to put the system into practice, wemust make identification of vehicles and their movement locus on the image we haveshot (Taking a photo of a vehicle is to identify and match any information concerningthe vehicle, its movement locus in case any illegal traffic behavior is committed).Therefore, to keep a track of vehicles becomes critically important. So far, there havebeen numerous research subjects home and abroad in this aspect. At present, theexistence of vehicle shadows exerts rather negative impact on computerized tracking.Therefore, how to eliminate the shadows of vehicles become the theme of the paper.This essay will solve this issue, the optimizing work concerning the calculationmethod of the present image cutting shadow is to be carried out, while the computer isanalyzing the image, this calculation method is being used to cut the shadow formedby the vehicle, locate the moving object in the scene accurately and obtain the correctinformation of the vehicle.This method fully takes advantage of the feature that the color saturation doesnot change, though the luminance of the image in the shadow changes, and it dividesthe image into several small images, compares the luminance and color saturation′schange of the corresponding small images in the former and latter images, lock thenon-shadow area, and then accurately locate the actual moving object.Analyzed from the result of the experiment, the effect of this calculation methodis obvious. It not only lowers the calculation pressure of the system, but also improvesthe identification efficiency of the system. And the vehicle′s accurate identificationalso provides a quite stable and authentic data source for other programs′development.
Keywords/Search Tags:smart traffic, illegal traffic behavior, vehicle identification, shadow
PDF Full Text Request
Related items