code

a lane dtection method using hough-transform

CNN中的手写识别LeNet5实现，用python，进行了中文注释(CNN handwritten recognition LeNet5 implementation, using python, carried out in Chinese Notes)

基于matlab的色差图像分割程序，可以实现复杂背景下的图像分割。(Matlab based on the color image segmentation procedure, the complexity can be achieved in the context of image segmentation.)

说明：  用c写的jpg编解码，包括jpeg2bmp和bmp2jpeg，可在vc6下调试。(With c written jpg codec, including jpeg2bmp and bmp2jpeg, may be under vc6 debugger.)

训练图片中检测到的目标，检测特征是基于梯度方向直方图(training target)

二值图像的Zhang快速并行细化算法，具有速度快、保持细化后曲线的连通性和无毛刺产生等优点。(Zhang fast parallel thinning algorithm)

读入人脸图像，根据图像的水平，垂直灰度积分情况，将人眼范围取出，背景可以较复杂。(Reading into the human face image, according to the image of horizontal, vertical gray points, which will remove the scope of the human eye, the background can be more complicated.)

基于DCT方法的数字水印论文5篇，对学习基于DCT方法的数字水印初学者会有帮助。可以了解基于DCT的水印基本方法。(DCT-based method of digital watermarking papers 5, to learn methods of digital watermarking-based DCT would be helpful for beginners. DCT-based watermark can learn the basic method.)

利用ENVI/IDL，计算多幅图像的各个像元的平均值。并忽略空缺值(Use ENVI/IDL, calculate the average of each pixel of multiple images. And ignores the value of vacancies)

说明：  利用线性规划问题求解稀疏恢复问题的基追踪（basis pursuit，BP）算法，研究了在含有脉冲噪声干扰的原始信号环境下对经典基追踪算法的改进。(basis pursuit (BP) algorithm)