Division Method Of Tire Surface Area

Surface areas of car tire record information of various parameters of the tires. If theidentifications of surface areas of car tires occur typographical errors, which will bemisleading to consumers and may cause a lot of accidents. Therefore, in order to ensure theprinting identifications of surface areas of car tires to meet the design requirements before thetires are used, we need to test the printing identifications of surface areas of car tires whichhave been produced. In most tire manufacturing company, the identified regions of the tiresurface are detected manually. Because the time of artificial detection is relatively long, theaccuracy is not high and is prone to error detection, omissive detection, etc., it is difficult toadapt to the needs of large scale production in tire manufacturing company, so automaticdetection technology of identified areas of tire surface has become a hard problem which tireproduction manufacturers urgently need to overcome. This paper design a system which partsidentified areas of tire surface and the matching and division of identified areas are the coreissues of the paper.This paper examines the division method of tire surface areas under the standard areafeature library. Firstly, we analyses the design program of the standard area feature library andgive the overall framework and establishment process of the standard area feature library;Then, using the SURF algorithm extracts the feature points in the standard areas we haddivided in advance, which completes the establishment of the standard area feature library;Next, the tire surface regional image is preprocessed, including median filtering, backgroundremoval and positioning, to eliminate the invalid regions on the image, which makes the wholesystem effectively enhance the computing speed and reduce redundant calculations, then againusing SURF algorithm extracts the feature points on the image which has been removedineffective areas; finally, using Euclidean distance method matches the feature points of thestandard area image and the image which will be matched. For mismatching problem offeature points matching, using random sampling consensus algorithm removes the point ofpoor quality as an external point. The division of tire surface identified areas is completed.This paper discusses implementation process of tire surface area matching system whichhas been developed by using OpenCV at the end of this article, and through a lot of picturestests the algorithm and implementation system and gives results.