Description: 针对可见光与红外图像的特点和难点，提出了可见光与红外图像配准与融合中的关键技术，即:
使用新型的基于一维最大类间方差和最大连通性测量的图像分割方法对源图像进行分割来更好地实行图像粗
配准 使用新型的特征点提取方法，特征点的匹配及误匹配的消除来更好地实行图像精配准 采用新型的基
于区域的树状小波活性测度计算来实现树状小波图像融合 利用自生成神经网络来实现模栩图像融合.
-For visible light and infrared images of the characteristics and difficulties, a visual and infrared image registration and integration of key technologies, namely: the use of new one-dimensional based on the largest variance between-class connectivity and the largest measuring method of image segmentation source image segmentation to better implementation of coarse image registration the use of a new type of feature point extraction method, feature points of the match and the elimination of false matches to better the implementation of image fine registration the use of new tree-based wavelet calculated to measure the achievement of the activity tree wavelet image fusion the use of self-generating neural network to achieve image fusion Xu mode. Platform: |
Size: 482304 |
Author:媛媛 |
Hits:
Description: 由于选用部分Hausdorf距离进行图像匹配时不能满足较大尺寸图像匹配实时性要求，利用了多级分层匹配技术加速图像匹配，本文提出了多级实时图像匹配算法，该算法以部分Hausdorf距离为图像匹配相似性度量，结合小波变换多分辨率思想提取多尺度图像特征，实现了由粗到精多级分层实时匹配。-As part of the Hausdorf distance used for image match the larger size can not meet the requirements of real-time image matching using a multi-layered image matching technology to speed up matching, this paper proposes multi-level real-time image matching algorithm for the image to part of the Hausdorf distance matching similarity measure, combined with ideological and extract multi-resolution wavelet transform multi-scale image features, achieved by the coarse to fine multi-layered real-time match. Platform: |
Size: 265216 |
Author:wenping |
Hits:
Description: 提出一种多尺度方向(multi-scale orientation,简称 MSO)特征描述子用于静态图片中的人体目标检
测.MSO 特征由随机采样的图像方块组成,包含了粗特征集合与精特征集合.其中,粗特征是图像块的方向,而精特征
由 Gabor 小波幅值响应竞争获得.对于两种特征,分别采用贪心算法进行选择,并使用级联 Adaboost 算法及 SVM 训
练检测模型.基于粗特征的 Adaboost 分类器能够保证高的检测速度,而基于精特征的 SVM 分类器则保证了检测精
度.另外,通过 MSO 特征块的平移,使得所提算法能够检测多视角的人体.通过对于 MSO 特征块的装配,使得算法能
够检测人群中相互遮挡的人体目标.在INRIA公共测试集合及SDL多视角测试集合上的实验结果表明,算法具有对视角与遮挡的鲁棒性和较高的检测速度. -The multi-scale orientation (MSO) features for pedestrian detection in still images are put forwarded in
this paper. Extracted on randomly sampled square image blocks (units), MSO features are made up of coarse and
fine features, which are calculated with a unit gradient and the Gabor wavelet magnitudes respectively. Greedy
methods are employed respectively to select the features. Furthermore, the selected features are inputted into a
cascade classifier with Adaboost and SVM for classification. In addition, the spatial location of MSO units can be
shifted, are used to the handle multi-view problem and assembled therefore, the occluded features are completed
with average features of training positives, given an occlusion model, which enable the proposed approach to work
in crowd scenes. Experimental results on INRIA testset and SDL multi-view testset report the state-of-arts results on
INRIA include it is 12.4 times the faster than SVM+HOG method. Platform: |
Size: 1868800 |
Author:尹世荣 |
Hits:
Description: 图像修复是指恢复图像中破损区域的颜色信息或者去除图像中的多余物体。针对视频图像损坏特征中较复杂的水
平与垂直条带，利用视频图像时间连续性的特点，提出基于小波分析理论的修复方法。实验结果表明，该系统能够较好地恢复
视频图像和有效去除文字。
-Image inpainting refers to restore the color information of the damaged region in the image or remove the
unnecessary object in the image. A method is introduced to compute the main direction of the texture and complete
the image based on wavelet transform. The core of the algorithm relies on using texture detection in a coarse-to-fine
scheme. Experimental results show that the algorithm can obtain the satisfied restoration video frames.
【Key words】video frame restoration wavelet transform video processing text extraction Platform: |
Size: 5686272 |
Author:孙红娟 |
Hits:
Search - coarse to fine wavelet