Image Feature Extraction Based on Micro imaging Technology and Data Mining
With the development of science and technology and the mass production and communication of image information, the application of automatic recognition and classification of target image is more and more extensive, and the requirements of target recognition technology for practical application are higher and higher. Image recognition and classification technology generally includes several stages: image preprocessing, image feature segmentation, feature extraction and automatic classifier design, among which the content and method of feature extraction are particularly important. Feature extraction provides all the basis for the classifier and its characterization performance directly determines the efficiency and accuracy of the classification. In practical problems, feature extraction may be interfered by noise such as background, Angle of view, illumination and size, etc. How to extract feature parameters with good characterization performance and less interference from noise is an important issue of applied research.