DOI: 10.1016/J.CHEMOLAB.2006.08.012
关键词: Measure (data warehouse) 、 Multiple discriminant analysis 、 Wavelet 、 Mathematics 、 Pattern recognition 、 Transformation (function) 、 Linear discriminant analysis 、 Blood smear 、 Feature vector 、 Artificial intelligence 、 Computer vision
摘要: Methods for processing and analysis of medical images are observed. A method recognition classification blood cells, based on the black-box model, is proposed. Two approaches (wavelet transformation angle measure technique) used construction feature vectors PLS discriminant discrimination classification. The tested low-quality smears shows high level cell type recognition.