搜索资源列表
OpenCV-Mean-Shift-demo
- 基于OpenCV实现的图像特征区域跟踪程序，基于类Mean-Shift实现的颜色直方图概率匹配算法，可同时跟踪6个特定目标区域。-OpenCV implementation of image features based on regional tracking program, based on class Mean-Shift probability of achieving the color histogram matching algorithm, which can track a
MultiVedio
- 使用OpenCV编写的多目标跟踪的视频图像处理，用鼠标拖动跟踪区域-The video image processing, multi-target tracking using OpenCV prepared dragging with the mouse tracking region
kalmanPcamshift
- 基于视频的跟踪，结合卡尔曼和camshift的目标跟踪。由鼠标选中目标区域（包含了鼠标选中程序），kalman和camshift预测质心，从而达到目标跟踪。-Based opencv, combined with the target tracking Kalman and camshift. By the mouse to select the target area, kalman and centroid camshift forecast to achieve target tracki
lunkuo
- 本代码基于opencv2.2，轮廓检测，输入图像，进行边缘检测并将图像区分为背景和目标，二值化，用轮廓把目标区域显示出来，还可得到目标区域内的面积，即目标的大小-The code is based on opencv2.2, contour detection, the input image, edge detection and image area is divided into the background and objectives, the two values, the targ
moving-object-detection-method-
- 针对视频监控系统的运动目标检测部分,介绍了以OpenCV为平台,以背景减法与帧差法为基础的检测方法。它以背景减法建立和更新背景模型,利用感兴趣区域的设置提高实时性,引入修正的大津法自适应阈值改善二值化效果,并以帧差法为补充降低虚报率及光照变化的影响。实验表明,该方法可以快速有效地对运动目标实施报警。-OpenCV-based moving object detection method and application of some of the articles I read, I feel
cvwork_correlation
- 使用vc++6.0和opencv1.0实现的模板与目标图片的匹配，用模板图像通过依次与图像的重叠区域进行相似度比较，将相似度最大的区域作为匹配目标区域。得出最后的匹配结果-Using vc 6.0 and opencv1.0 achieve the target image matches the template, using the template image followed by image overlap region with similarity comparison, simi
VideoDetect
- 基于OpenCV的简易目标检测跟踪程序，能够实现对每帧视频的连续编号截图，可以实现键盘对视频播放的控制，能对每一视频帧生成包含运动目标区域的前景图像，采用某种策略实现对运动目标的跟踪，并用矩形框+编号将运动目标标出。-A simple target detection program base on OpenCV , which can numbered consecutively for each frame of video screenshots, keyboard control the
test2
- 检测区域红色方框像素并提取中心点，并可以动态跟踪目标达到大约15帧秒的处理速度。-The red square pixel of the detection area and extract the center point, and can dynamically track the target to achieve the processing speed of about 15 seconds.
ParticleFilter
- 颜色直方图粒子滤波器，并给出了英文参考文献； 运行方式：在前景窗口，按p键停止，在目标区域点击鼠标，让程序自动识别出目标轮廓，再次按p键，即可跟踪-The color histogram particle filter, given English references operation modes: in the foreground window, press p stop clicking the mouse in the target area, allowing the pro
FindMoving
- 由运动目标检测的两种基本方法----帧间差分法和背景差分法,借助于OpenCV技术，在Visual C++ 6.0编程环境下开发了运动目标检测系统。该系统首先对不同途径采集的视频图像序列进行相关的预处理之后，分别采用不同检测算法检测出图像序列中的变化区域，最后用形态学滤波和连通性分析对变化区域进行后处理，从而将视频图像序列中的运动目标比较可靠地检测出来。-The two basic methods of moving target detection---- inter-frame diffe
Moving-target-detection--opencv
- 运动目标自动跟踪检测，选择一块区域，根据一些色差值等，检测相关点-Moving target tracking and detection automatically select an area, according to some color difference detection point
stronger
- 图像增强算法，对图像进行边缘提取，能够获得目标区域-failed to translate
contour
- 基于opencv的计算图像的边界区域，从而提取出目标区域，达到图像分割的目的。-Calculated based on the opencv image boundary region, thereby extracting the target area, to achieve the purpose of image segmentation.
LOT_Source
- 目标跟踪第t帧图像，目标特征，噪音参数，粒子窗口 1. 在第一帧选定跟踪区域，将图像由rgb转成hsv，计算目标的特征模型； 2. 初始化粒子，用超像素计算粒子的特征； 3. 通过emd计算粒子所在位置区域特征计算与目标特征的相似度； 4. 根据emd计算粒子的权重，归一化权重，新的目标位置是所有粒子的位置和权重乘积的和； 5. 建立新目标的特征模型，计算emd。 6. 更新噪音参数； 输出：目标位置，更新参数，新的粒子窗口 -Multiple Object Tr
Image-expansion-areas--corrosion-
- 主要用于空间视觉定位时，目标区域的形态学处理，达到后期处理要求-When space is mainly used for visual orientation, morphology of the target area, which will reach the post-processing requirements
Face-Detection-ren-lian-jian-ce
- 目标跟踪,利用相邻两帧的区域匹配从图像序列中建立目标链,跟踪目标从进入监视范围到驶离监视范围的整个过程-Target tracking, the use of two adjacent frames matching goal of establishing regional chain from an image sequence, tracking targets range from entering the monitor to monitor the whole process of
Meanshift
- 摄像头固定情况下的运动行人跟踪进行研究，Mean Shift算法，与目标区域灰度直方图结合，进行运动目标的检测跟踪-tracking study of the movement of pedestrians in case of a fixed camera introduces a better application of Mean Shift algorithm
HandDetector
- 这是一个人手识别程序，基于OpenCV，环境VisualStudio2013。 首先通过图片采集手势图像，而后进行减小噪声的滤波处理，根据肤色在YCrCb色彩空间中的自适应阈值对图像进行二值化处理，提取图像中所有的肤色以及类肤色像点，而后经过膨胀、腐蚀处理后，祛除小块的类肤色区域干扰，得到若干块面积较大的肤色区域；此时根据各个肤色区域轮廓特征进行甄选，获取目标手势区域，而后根据目标区域的特征进行识别，定位手心。-This is a staff recognition program, bas
test3
- VS2013+OPENCV2.4，OTSU阀值分割确定目标区域，sobel差分算子用来进行边缘检测，可知程序运行时间-VS2013+OPENCV2.4, OTSU segmentation threshold determining target region, sobel operator for performing differential edge detection, it can be seen running time
PeopleTracker-master
- 通过摄像头拍摄来监测目标区域，对目标区域的行人作实时检测-hrough the camera to monitor the target area, pedestrian for real-time detection of the target area