资源列表
MeanShift
- 为了提高经典的Mean Shift算法在复杂场景中的跟踪性能,提出了一种基于角点的目标表示方法。首 先,利用Harris角点检测算法提取表示目标主要特征的角点 其次,基于提取的角点,建立目标模型,将其嵌入 Mean Shift算法进行跟踪。该方法仅用少量的关键点表示目标,能够自动去除目标和背景中的次要特征,有效地 抑制背景成分对目标定位的影响,从而改进Mean Shift目标跟踪算法的性能。-To imp rove the performance of the classicalMe
b_back
- 主要用于校正图像的畸变，用matlab编写的，很实用-Mainly used to correct image distortion, written with matlab, very practical
K-MEANS
- 基于K-MEAN的图像分割,方便实用，对于图像处理的研究生很有参考价值的-watershed segmentation on matlab
a-level-set-inariable-distance
- 李春明提出的无需水平集初始化的思想是图像分割的一个重大突破，重庆大学的何传江教授对该算法进行了改进，里面是我的实现算法-Li Chunming s no need to initialize the idea of the level set image segmentation is a major breakthrough, Chongqing University, Professor He Chuanjiang to improve the algorithm
MSR
- MSR单尺度matlab代码，直接运行即可-MSR single scale matlab code can be run directly
Gaussian_Lowpass_Filter
- 二阶高斯低通滤波器，实现对图像的平滑处理，-Gaussian second-order low-pass filter, to achieve the smoothing of images,
matlab
- 汽车车牌数字识别（上个学期的数字图像处理的课程设计），能运行-Automobile license plate number identification (the last semester of digital image processing course design), can run
figure-of-merit-
- 评价图像边界的好坏，FOM值越趋于1质量越好。-figure of merit to objectively evaluate the performance of edge detectors
GraphCuts
- 基于图论的高效图像分割方法，比NCut算法的效率更高，效果更好-Based on Graph Theory and efficient image segmentation method, an algorithm NCut more efficient, better
LiveWireFinal
- 交互式图像分割的一种算法——LiveWire在VC++平台下基于opencv的实现。类似于Photoshop中吸铁石套索的功能。-Interactive image segmentation algorithm- LiveWire platform in VC++ opencv-based implementation. Similar to the magnet lasso feature in Photoshop.
MW_denoise
- 此程序是一维信号多小波去除噪声的实现程序。-1-D signal denoising by multi-wavelet and cycle-spinning.
lap_fusion
- matlab基于拉普拉斯金字塔分解的图像融合源程序,已经过测试，可以应用。-matlab based on the decomposition of the Laplacian pyramid image fusion source