Research And Implementation On Swine Body Identification Based On The Characteristic Of Image Texture And Color

In numerous image information technologies, pattern recognition is one of the most widely used techniques. It is widely applied in various fields such as automatic identification system, safety monitoring and video monitoring of product line, etc. Getting useful image features is the key step of pattern recognition, which includes contour extraction,texture feature and semantic extraction, etc. These are the hotspot and difficulty of recent research. Target object recognition is one of typical applications of pattern recognition. People have paid more and more attention on processing platform target object recognition and monitoring system.After consulting a large number of literature and analyzing the status quo of pattern recognition technology at home and abroad, this thesis have studied the pig-recognition in a real project. Based on pattern recognition technology, we designed the relevant algorithm using the texture and color feature. This algorithm includes getting and choosing the texture and color characteristics and classification, etc.We used two ways to get the texture feature of the object. They are multi-scale and multi-direction gray level co-occurrence matrix and multi-channel Gabor filters. Both of them could reflect the information in different directions and scales, which could fit for the way that human percept the images. In this paper, we have got the color feature thorough two different ways that is color model and texture feature in YCbCr color field. Texture and color feature are used as image features to classify images. We have used two ways in this paper, the first one is separate method that is separately get the texture feature in gray image and pure color characters in different color fields, then use the character-melt technology to put them in one character vector. There are color and texture information in one character vector. Though this way, we have composed two features in different fields together. The other way is using united approach that is in color field we received the texture feature like that is used in the gray image.This object based on an auto-system that is used in pig-breeding factory. Combined computer and digital image treating technology,this paper studied the auto-reorganization technology, gave out the relevant showed the experiment results. The result has shown that texture and color feature could effectively locate pig and give out the total number of them. The widely use of this method could highly improved the auto-recognize and count efficiency, improve labor productivity level and promot the pig breeding and automation management scale.