Research Of The Algorithm In Comparison For Footprints Based On LBP And Shape Context

Footprint trace is one of the important evidence can be found at the scene of the crime, has the characteristics of easy to extract.If footprint trace can be better used, policeman analysis of the case more realisticlly, and footprint will play an important role in getting the contection between law cases or detecing law cases. However, if only the footprint traces from one scene of the crime were analysised,only the quantily of people who committed the crime can be infered.It is obvious that the footprint trace didn’t be taken full advantage of. At present, the public security organs in order to make full use of the left footprints, has established a relatively perfect library of footprint images which were lelf in scene of crime. However, through the method of artificial to waste a lot of manpower and time. Therefore, through the computers and technology of image processing to realize matching for images of footprint and retrieve has become a practical new topic.This article uses LBP and Shape context to get image features and uses method of image processing to compare images and retrieval. When LBP is used as feature. using the LBP rotation invariant pattern to ensure the image alignment rotation invariance. The cascade method of multiple scales is applied, so that full use of the local information of the image. Using comparison intercept method to reduce influence from uncompleted images. When Shape context is used as feature, in order to get down the influence from uncompleted images, proposed a weighted shape context method. Application of binary chart optimal match method, the repeated use of feature points are avoided. By using the method of triangle screening to remove the false match point, improved the effect of image comparison and retrieval.All the modules of core algorithm are tested and realized by MATLAB. In order to improve the work efficiency, VC is used to design the interface. Using the interface of VC to call the modules in MATLAB can realize that paramaters were inputed in conviently and results were outputed in visually.