基于蚁群算法的图像边缘检测

图像仿射匹配的经典文章，适合深度学习仿射匹配的研究生(Image the affine matching the classic article, suitable depth learning graduate affine matching)

matlab数字图像处理的程序，里面有基本的处理程序和GUI界面来处理图像(Matlab digital image processing program, which has a basic processing program and an image processing GUI interface)

改进后的非线性各向异性扩散用于图像中的保边去噪，增加了二阶导数，是滤波效果更好(Improved nonlinear anisotropic diffusion for edge preserving image denoising, an increase of second derivative, is a better filtering effect)

说明：  用数学形态学滤波进行图像处理或者信号降噪，（包括膨胀和腐蚀）(Image processing or signal noise reduction using mathematical morphological filtering (including expansion and corrosion))

K-Means算法,不要求建立模型之后对结果进行新的预测，没有相应的标签，只是根据数据的特征对数据进行聚类。主成分分析降维对数据进行可视化操作,对features进行降维.(K-Means algorithm does not require the establishment of the model after the new prediction of the results, there is no corresponding tag, but only on the characteristics of data clustering data. The principal component analysis reduces the dimension, carries on the visualization operation to the data, reduces the dimension to the features.)

说明：  区域生长法的图像分割方法，效果分厂不错，值得借鉴！(Region growing image segmentation method, the effect of a good factory, worth learning!)

这是一个数字模糊处理的算法论文!很好很经典(Number obfuscation algorithms! Good classic)

图像检测的算法，Faster R-CNN算法，先对整张图像进行卷积计算，然后通过感兴趣区域池化层(RoI Pooling Layer)将选择性搜索算法推荐出来的候选区域和卷积网络计算出的特征映射图进行融合，得到候选区域对应的特征矢量，这种共享卷积计算的操作极大地减少了卷积计算的次数。而且这些特征矢量的维度统一，方便后续的分类工作。通过感兴趣区域池化层处理卷积特征，并将得到的特征送往两个并行计算任务进行训练，分类和定位回归。通过这些方法和改进的框架，Fast R-CNN 用更短的训练和测试时长，取得了比 R-CNN 更好的效果(Faster R-CNN algorithm first convolutes the whole image, then fuses the candidate regions recommended by the selective search algorithm and the feature mapping maps calculated by the convolution network through the RoI Pooling Layer to get the corresponding feature vectors of the candidate regions, which greatly reduces the number of convolution calculations. Moreover, the dimension of these feature vectors is unified, which facilitates the subsequent classification work. The convolution feature is processed by the pooling layer of the region of interest, and the obtained feature is sent to two parallel computing tasks for training, classification and positioning regression. Through these methods and improved framework, Fast R-CNN uses shorter training and testing time and achieves better results than R-CNN.)

说明：  关于GLSL的点光源，载入模型，纹理。。。。。。。。。。(light engine code OpenGL.)

自写的一种基于局部对比增强的GVFsnake图像分割方法，参考了xu的gvf，但本方法分割效果更好。(Since the writing of a local contrast enhancement based on image segmentation method of GVFsnake reference xu the gvf, but this method is better segmentation.)