This paper presents a text line segmentation method for printed or handwritten historical Arabic documents. Documents are first classified into 2 classes using a K-means scheme. These classes correspond to document complexity (easy or not eas

对4幅字符图像进行字符分割（基本要求）以及字符识别。或者对一段视频中的每帧图像进行字符分割与识别。-Four character image segmentation of the characters (basic requirements) and character recognition. Each frame in a video character segmentation and recognition.

基于树莓派B+版本的车牌识别设计，辨识部分以树莓派3B+/3B为核心处理器，基于Linux平台，通过python3.5和OpenCV来对含有车牌信息的图片，进行高斯去噪、灰度化、边缘检测、二值化、闭操作、字符分割、字符识别匹配等一系列图像处理操作最终实现车牌识别这个系统。(Based on the design of license plate recognition of raspberry pie B+version, the recognition part takes raspberry