基于BP神经网络对印刷体字符图象进行识别,经过前期处理,获得64*64的二值图象,其二值图象数据作为神经网络的输入.-BP neural network based on the characters of the print images to identify, after pre-treatment, access to 64* 64 binary image, and the second value of image data as the neural network input.
经过前期的数字图像处理以及检索的技术研究之后,可以根据图像的数据格式、显示方式等特点决定采用哪种框架。-digital image processing and retrieval of technical studies, according to the image data formats, displays the characteristics determine what framework.