搜索资源列表
ssim_index
- Zhou Wang proposed a new idea of Structural Similarity Index (SSIM )for image quality measurement based on the assumption that the HVS is highly adapted to extract structural information from the viewing field. The Mean Structural Similarity Index (M
meanshift
- 实现了基于mean-shift的图像检索，实现了比较两图像的相似度，选择最相近的图片-To achieve mean-shift based image retrieval, to realize the similarity of two images, the selection of the most similar image
image_fusion_proforma_evalu_quality
- 这是从网上整理出来的图像融合评价标准，总共有13项性能指标。包括平均梯度，边缘强度，信息熵，灰度均值，标准差(均方差MSE)，均方根误差，峰值信噪比（psnr），空间频率(sf)，图像清晰度，互信息(mi)，结构相似性(ssim)，交叉熵(cross entropy)，相对标准差。大家一起交流吧~-This is sorted out from the online image fusion evaluation criteria, there are a total of 13 perform
MSRM_1
- 程序代码说明：本程序是应用于图像分割（区域合并）基于最大相似性区域合并的交互式图像分割 本代码绝对可用-Code Descr iption: This procedure is applied to image segmentation (region merging) region merging based on similarity of the largest interactive image segmentation of the code is absolutely free
10
- 图像变换代码 图像配准 包括点映射相似性测度的选择等-Image transform code including image registration point mapping similarity measure of choice
program
- 本程序用于对工件图像的预处理，边缘提取等。将改进遗传算法和Hausdorff距离结合用于工件识别。采用canny边缘为匹配特征，将修正的Hausdorff距离作为相似性度量准则，并应用遗传算法进行最佳匹配的快速搜索，成功实现了目标物体的匹配识别-This procedure used for the workpiece image preprocessing, edge extraction and so on. Will improve the genetic algorithm and Ha
Tracking
- 提出一种新的目标表示和定位方法,该方法是非刚体跟踪的核心技术.利用均质空间掩膜规范基于特征直方图的目标表示,该掩膜引入了适合于梯度优化的空间平滑相似函数,所以可以将目标定位问题转换为局部极大值求解问题.我们利用从Bhattacharyya系数倒出的规则作为相似度量,利用mean shift procedure完成优化求解.在给出的测试用例中, 本文方法成功解决了相机移动,阴影,以及其他的图象噪声干扰.文章对运动滤波和数据关联技术的集成也进行了讨论.-A new objective and pos
facerecognizeonPCA
- 像的许多个空间中的一个。不管子空间的具体形式如何，这种方法用于图像识别的基本思想 都是一样的，首先选择一个合适的子空间，图像将被投影到这个子空间上，然后利用对图像 的这种投影间的某种度量来确定图像间的相似度，最常见的就是各种距离度量。-Like much of a space. Regardless of the specific sub-space form, this method has been applied to image recognition are the same
paizhao
- 通过CCD拍照后分析所拍图象的坐标以及相似度。-Through the CCD camera image after analyzing the film as well as the similarity of the coordinates.
ImgComparator
- 实现了基于Gabor滤波器的图像比较，比较两张图片的结构相似性-Gabor filter based on the realization of the image compares the two pictures of the structure similarity
oula
- 基于欧拉-熵向量的图像检索,相似度可用马氏距离-Based on Euler- entropy vector image retrieval, similarity can be used Mahalanobis distance
browser_browser
- The goal is to provide user with a systematic navigation tool for image database access. In this system, user can use an example image to retrieve similar images or use a 3-D location to retrieve a snapshot of the similarity pyramid of the databa
shadow5
- 分块的阴影检测方法：将每帧图像进行合理的分块,并且采用基于梯度的方法对运动区域边缘的小块进行合并。对每个小块根据阴影区域和对应的背景区域之间具有较强的结构相似性和色度近似性的特点进行阴影检测。 -The shadow of block detection method: The image of each frame block reasonable and based on the gradient method of movement of the edge of the small r
xiangsidu
- 遥感图像的多光谱目标识别，相似度方法实现分类-The multi-spectral remote sensing image object recognition, similarity classification method
090211
- 细化参数的对数极坐标变换图像纹理特征提取算法提出一种细化参数的对数极坐标变换图像纹理特征提取算法,可以有效地消除旋转、缩放和平移等几何形 变的影响.在特征提取过程中,通过自相关图像消除平移的影响,引入细化参数的对数极坐标变换消除旋转和伸缩 的影响.在图像检索过程中,将不完全树型小波变换所得的特征矢量,通过欧式距离来度量图像之间的相似度.实 验表明,本算法对发生几何形变的纹理图像检索平均正确率达81.05 左右,较之传统算法能取得更好的检索 效果.-This paper p rop
dlez
- 基于HSI色彩空间的道路区域去除 通过对RGB空间的色彩变换，根据颜色的不相似度量，将道路图像中的色彩区域去除，由于阈值设置的问题效果有时不是太好，还需要进一步完善。-HSI color space based on the removal of the road area Through the RGB color space conversion, according to the color similarity measure is not, the road to remov
ChildFrm
- The first scheme is based on the spatial locality of feature vectors corresponding to similar images. Learning is effected by modifying the query vector to incorporate the positive examples. The second scheme is based on “distorting” our view of
key_frame_get_source
- 实现“基于互信息量”进行关键帧提取的具体算法，该算法以帧间的互信息量来衡量两帧图像内容的相似性，可以对具体镜头按其特点自动提取出一或多个关键帧。-To achieve " Based on mutual information" to a specific key-frame extraction algorithm in order to frames to measure the two mutual information similarity of image cont
InteractiveImageSegmentationbasedonMergingRegion.r
- 基于区域融合的半监督的图像分割算法。首先在背景和前景手动设置初始分割标记，在迭代过程中不断通过区域融合操作获得最大相似度的区域，从而实现目标分割。-Regional integration based on semi-supervised image segmentation. First of all, in the background and foreground segmentation manually set the initial marking, in the iteration
boxcount
- A set (e.g. an image) is called "fractal" if it displays self-similarity: it can be split into parts, each of which is (at least approximately) a reduced-size copy of the whole. A possible characterisation of a fractal set is provided by the "box-