基于拉haar小波的图像增强，与传统直方图均衡化相比较，效果更好。...

人脸识别的源程序，尝试一下如何上传，下次再传好的-Face recognition source code, try how to upload, send the next best

Character Recognition Using Neural Networks Steps to use this GUI. 1. Open the GUI figure, run it. (accept the matlab to change its directory to new location where the file is stored) 2. First we need to teach Character to computer. For this type the Character in the textbox space provided and press "TEACH". 3. You can save all the taught data. 4. For retrival, click start. Note: Paint Brush software of microsoft opens for entering the image. So, we need to save the image after editing. For best results teach each Character not less than ten times.

03 VC图像分割与边缘检测源代码，用不同方法实现图像分割-03 VC image segmentation and Edge Detection source code, using different methods of image segmentation

传统的单粒子重建方法使用傅立叶或delta 函数基础来表示粒子密度图。我们提出了一个更灵活的算法，可以根据数据适应性地选择基础。由于基础适合数据，跟使用固定基础重建相比，分辨率和讯噪比(SNR) 都得到提高。而且，该算法自动遮盖粒子，使得粒子从背景中分离出来，无需在精制循环中使用专门滤波器或者蒙层。该算法是基于贝叶斯定理最大后验框架中表示，使用有效优化算法获得最大化。评估使用模拟和实际的低温电子显微镜数据，显示分辨率和 SNR 改进以及从背景中有效遮挡粒子。

通用的去验证码中干扰线的资料,使用最短路径和方向偏移算法实现-A common code to verify the information in the interference lines, use the shortest path and direction of the offset algorithm

DSP汇编进行简单的图形处理，例如平滑与锐化等-Using DSP s assembly language to do Image Processing such as smoothing and acuting

标准测试图像下载-standard test images downloaded

椭圆的有效的扫描转换-elliptical effective scan conversion

用数据拟和曲线的工具，利用了三次样条插值等方法。-data used to be and curve tools, the use of a cubic spline interpolation methods.

多线程图形特效，可以预览到多线程播放图片的许多效果，非常方便。-Multi-threaded graphics effects, you can preview images to a multi-threaded playback of many effects, very convenient.