| Application of wavelet analysis on automatic impurity recognition on image of gizzard packed in the aluminum foil detected by X-ray technology.With the improvement of life quality, people pay more attention to the food safety than before. More and more specialists are focusing their researches on the methods of detecting and identifying foreign objects in food, one of which is X-ray inspection system.With the support from Science and Technology Project of Jiang Su province, this paper discuses application of wavelet analysis and Invariant Movements on automatic impurity recognition on image of gizzard packed in the aluminum foil detected by X-ray technology.In this paper, the characters of X-ray inspection system are described briefly, then the methods of image filtering and segmenting, template matching and identifying are researched in detail. Firstly, after comparing the results of image filtering based on wavelets named bior3.7, db1 and syms4, the first wavelet is chosen to reduce the image noise. Secondly, Bimodal law, Iterative law and OSTU law are utilized to segment the image templates, which are constructed to match with the source image. Finally, the invariant moments were used to extract the features of the image templates that are transformed by wavelet. It achieves identifying the impurity in the gizzard images after analyzing similarity through the Euclidean distance between two images' normalized moment vectors in the experiments.A man-machine interface is designed based on the algorithm in this paper, which can reduce the image noise, extract the invariant moments of wavelet matrix and identify the impurity. According to the experimental results, this algorithm has demonstrated good performance in recognition of various impurities in gizzard by eliminate the background intense effect to the template matching. It can work accurately and quickly, which is worthy of spreading and applying in produce. |
PDF Full Text Request