In this paper, the image shape feature extracted by the PCA method can meet the needs of the recognition layer well.
在该文的形状识别系统中，用P CA法提取图像的形状特征，能够较好地满足识别层的输入要求。
The 'lines' need not be actual lines - they could be the shape of a path, a line of trees, a fence, river or any other feature in an image.
所谓的线并不是条实际存在的线——它可以是一条小路、一排树、一排围栏、一条河或者其他任何在图中出现的东西。
The contents in the paper include acoustic image preprocessing, feature extraction of texture and shape, and classifier design.
内容涉及声图像的预处理、纹理和形状特征的提取，以及分类器的设计等。
应用推荐