糜状食品物料微结构图象特征的小波分析

Wavelet Analysis on the Feature of Microstructure Image of Minced Food Material

由于糜状食品物料微观结构和流变特性密切相关 ,所以正确地识别和提取糜状食品微结构图象的特征 ,是建立流变特性计算模型的关键步骤 .本文基于多分辨率分析方法 ,对不同微观结构的糜状食品图象进行小波变换 ,研究了图象多尺度特征 .同时对变换后的小波系数进行统计分析 ,探讨了不同微结构图象小波变换系数统计量的区别及变化特点 ,并通过计算矩阵特征值 ,对建立流变特性计算模型的统计参数进行降维处理 .

Because of close relationship between the microstructure of minced food material and its rheological behaviour, correctly recognizing and extracting the characters of its microstructure image is one of the most important subjects to model the rheological behaviour. Images of various minced food material microstructures were wavelet transformed in this paper to study the multiscale features of the image. Simultaneously, the wavelet coefficient transformed was analyzed statistically to discuss the differences of their statistic data. In the end, the statistic parameters for rheological behaviour modeling were predigested by calculating the eigenvalue of the matrix.