Texture_sys

空间金字塔，通过对图片不断的进行层次划分，对每一层的图片信息进行整理用于图片分类等(Beyond Bags of Features: Spatial Pyramid Matching for Recognizing Natural Scene Categories )

改进的数学形态学小目标检测，用于红外图像中的点目标检测，比传统的Top-Hat变换检测精度高(Small target detection improved mathematical morphology for point target detection in infrared images, higher detection accuracy than traditional Top-Hat transform)

自行编写的逆投影变换，可实现图像坐标系到世界坐标系的转换，从实现鸟瞰变换。(the IPM ,which is completed by me, is able to achieve the transformation between the image coordinates and world coordinates )

通过计算机usb摄像头进行图像采集，然后实时处理，跟踪其中运动物体，并进行计数，采用中值滤波法得到背景，并根据需要进行更新，视频处理采用了动态差分与前景差分结合的方法，效果比较理想，但无法解决人物重叠的问题(Usb camera through the computer for image acquisition, and real-time processing and tracking moving objects in which, and the count median filter by using the background and update as needed, video processing using a dynamic difference with the prospect of differential combination of methods, results ideal, but can not solve the problem of overlapping characters)

运用matlab提取视频，显示出视频图像的图像特征，并以柱状图形式显示(Use matlab extract video shows the image feature of the video image, and is displayed in histogram form)

说明：  注：运行前自己准备图像，修改I = imread('')单引号中的路径 1. 给一张灰度图添分别添加高斯噪声和椒盐噪声。然后分别进行均值滤波和中值滤波，并把处理后的图片保存。 2. 图像频域处理，对图像进行傅里叶变换，进行高通滤波，显示处理结果。 3. 图像平移和旋转 4. 分别用OSTU阈值法和全局阈值法分割图像，然后进行腐蚀和膨胀处理，显示处理结果。(Note: Prepare the image by yourself before running, modify the path in single quotes of I = imread('')1. Add Gaussian noise and salt and pepper noise to a grayscale image. Then perform mean filtering and median filtering respectively, and save the processed pictures.2. Image frequency domain processing, Fourier transform of the image, high-pass filter, and display the processing result.3. Image translation and rotation4. Separate the image with OSTU threshold method and global threshold method, and then perform corrosion and dilation processing, and display the processing results.)

基于双目的深度图获取,根据图像合成深度图左右视点( depth map based on binocular)

高精度亚像素全局运动估计的上采样梯度互相关算法(Precision subpixel global motion estimation on sampling the gradient of cross-correlation algorithm)

SPCA The SPCA algorithm for computing sparse principal components.

IDL关于图像感兴趣图像区域进行编辑并将结果输出，多ROI区域DN值信息比较(IDL about the region of the image interested in editing and exporting the results, and comparing the DN value information of the multiple ROI regions)