Unit1

DensePose用深度学习把2D图像坐标映射到3D人体表面上，再加上以每秒多帧的速度处理密集坐标，最后实现动态人物的精确定位和姿态估计。该技术集目标检测、姿态估计、目标部分/实例分割等多种计算机视觉任务于一身的一个综合问题。(DensePost maps 2D image coordinates to 3D human body surface by in-depth learning, and processes dense coordinates at the speed of multiple frames per second. Finally, it realizes precise positioning and attitude estimation of dynamic characters. This technology integrates many kinds of computer vision tasks, such as target detection, attitude estimation, target part/instance segmentation, etc.)

EZW图片压缩，利用EZW原理在matlab里面编程实现图片压缩(EZW-compression of picture)

直方图特征的计算，包括方差，峰度，熵值等。(Histogram of the characteristics of the calculation, including the variance, kurtosis, entropy and so on.)

说明：  数字图像处理 小波处理 ，去噪音，比基于傅里叶变换的去噪方法好。 含噪信号经过预处理，然后利用小波变换把信号分解到各尺度中，在每一尺度下把属于噪声的小波系数去掉，保留并增强属于信号的小波系数，最后再经过小波逆变换恢复检测信号。(The noisy signal is pre-processed, then the signal is decomposed into different scales by wavelet transform. The wavelet coefficients belonging to noise are removed at each scale, and the wavelet coefficients belonging to the signal are retained and enhanced. Finally, the detected signal is restored by inverse wavelet transform. It is better than the denoising method based on Fourier transform.)

自己总结的JPG、GIF、BMP、PNG、WBMP图片文件格式文档以及相关参考标准与SPEC(their lessons JPG, GIF, BMP, PNG, WBMP picture files and related documentation and reference standard SPEC)

一个读取DXF文件的工具，有示例，有源代码，对某些需要与Autocad进行图形学接口编程的朋友有十分重要的工具，我可以千般辛苦找来的。(a read DXF files tools, example, the active code, certain needs and Autocad graphics interface for programming and friends very important tool, I can offer all kinds of hard to find.)

基于支持向量机方法的图像分割与目标分类.介绍了支持向量机用于车辆图像分类。以及用于单目标图像的分割和多目标图像的分割，介绍的较为详细。(Method of image segmentation based on support vector machine (SVM) and the target classification. This paper introduces the support vector machine (SVM) is used to vehicle image classification. And for single target image segmentation and target image segmentation, introduced detailed.)

提取视频中的图片，每一帧的信息，把视频分为图片信息(get picture of video)

说明：  平滑滤波，空域滤波与变换域滤波比较，从中说明其中原理(The principle of smoothing filtering is explained by comparing spatial filtering with transform domain filtering)

英文文本的相似度计算，分别从词形、词序、词义等方面进行权重计算，得到相似度结果(English text similarity calculation were re-calculated from the word form, word order, meaning, etc. right, the similarity results)