nmfsc

A Low Complexity Motion Compensated Frame Interpolation Method

说明：  人工智能 caffe 用与ssd yolo(Artificial intelligence SSD Yolo)

ostu二值化算法vc实现，基于MIL图像处理软件包将图像解析，通过ostu求取二值化阈值，实现二值化算法(ostu binarization algorithm to achieve vc, MIL image processing software package based on the image resolution, strike through ostu binarization threshold binarization algorithm to achieve)

对图像矩阵进行zigzag扫描,从而达到对图像纹理（灰度值）置乱的效果。( scanning of the image matrix by Zigzag，To achieve the scrambling effect of the image texture (gray value).)

自写M函数，做灰度图像的灰度级数及灰度范围的自定义设置。例如：将[0 255]的uint8图像变换成8级灰度的灰度范围为[0 7]图像或者[11.5 25.8]图像。 输出图像为double类，用户可将图像归一化后自行转换为所需类型。(This is a M function written by self, it can be used to rebuild a image with a given number of gray-level and a given gray-scale. Output image is double class.)

基于二维直方图的最大模糊熵阈值图像分割，相比于一维最大模糊熵，分割效果更好。隶属度函数采用S函数(Based on two-dimensional histogram maximum fuzzy entropy threshold image segmentation, compared to one-dimensional maximum fuzzy entropy segmentation better)

说明：  均值漂移聚类的源程序,用mean shift 进行聚类，以及绘图显示，很有价值(mean shift based clustering code, using matlab)

视频水印的matlab源码，文件较多，分为六个压缩包，如果您作相关方向的研究，可用作参考(Video Watermarking Matlab source, the paper more, divided into six compressed, If you made related to the direction of research, can be used as a reference)

说明：  压缩感知稀疏恢复中使用的OMP算法，matlab例程(OMP algorithm and MATLAB routine used in compressed sensing sparse recovery)

【实例简介】本设计为基于MATLAB的火焰烟雾火灾检测系统。传统的采用颜色的方法，误识别大，局限性强。结合火焰是实时动态跳跃的，采用面积增长率，角点和圆形度三个维度相结合的方式判断是否有火焰。该设计测试对象为视频，通过下一帧和上一帧的差异发现是否有火情，并可发出语音报警。本设计带有一个人机交互式GUI界面，界面友好。是个不错的bishe选题。