Font Size: a A A

The Research Of Vehicle Sign Recognition

Posted on:2008-02-05Degree:MasterType:Thesis
Country:ChinaCandidate:H W YuFull Text:PDF
GTID:2178360248452212Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
The thesis mainly carried on researching for vehicle sign location and recognition. It locates and recognizes the sign of vehicle, by utilization of optimized recognition algorithms and information fusion techniques, from the image of moving vehicles variant feature of vehicle in real-time, such as vehicle-license-plate and vehicle-sign.Location and recognition about sign of vehicle are difficult problem because of it's fixed attribute, such as small, similar each other, influenced by its size and the light, and different in size and shape . This thesis solves these problems by two aspects. The one is the researching the location of sign, the other one is the recognition. Furthermore vehicle sign location is more important key among the best, for directly influencing the recognition rate by location result. Recognizing the species of sign is the goal of recognition.First, all kinds of location methods from other documents are systematically analyzed, and some key problems on vehicle sign location are summed up. Second , for these problem , by full utilize of the location of vehicle-license-plate, and its correlation of vehicle-sign and interferential object, a vehicle sign location method is presented in the thesis. Begin, the vehicle radiator image, whose position is found by locating the vehicle license, is disposed. Then, based on the color and shape of radiator field, vehicle sign located is by OTSU dignitary and some image processing methods. This mainly method is mathematical morphology and projection. Last, by utilize of moment invariant features, the paper recognizes the vehicle sign. Moment invariant applies to the vehicle sign for it's feature, which relevant with the object shape, and irrelevant with the scale, translation and rotating invariant.Based on pattern recognition and image processing techniques, we studied such topics vehicle sign location and recognition. Experiment results show that, the technique presented in this paper is reasonable, and have theoretical and practical values.
Keywords/Search Tags:Vehicle Sign Location, Image Segmentation, Mathematical Morphology, Projection, Moment Invariant
PDF Full Text Request
Related items