根据CSLBP(center-symmetric local binary pattern)和Uniform LBP(local binary pattern)特征描述行人局部纹理互补性的特点,提出将二者级联的组合特征用于行人检测:基于灰度图像的纹理特征(hybrid local binary pattern,HLBP)和基于...根据CSLBP(center-symmetric local binary pattern)和Uniform LBP(local binary pattern)特征描述行人局部纹理互补性的特点,提出将二者级联的组合特征用于行人检测:基于灰度图像的纹理特征(hybrid local binary pattern,HLBP)和基于颜色空间的纹理特征(color based hybrid local binary pattern,CHLBP)。实验结果表明,当FPPW=10–4时,HLBP特征的检测率为93.96%,与Uniform LBP和CSLBP特征相比分别提高3.46%和9.68%,基于颜色空间L′C′C′与HIKSVM分类器结合时的检测率高达98.58%。与传统的纹理特征检测方法相比,该特征提高了行人检测精度,降低了误检率,检测性能得到较大幅度的提升。展开更多
In this paper, we present an efficient approach for unsupervised segmentation of natural and textural images based on the extraction of image features and a fast active contour segmentation model. We address the probl...In this paper, we present an efficient approach for unsupervised segmentation of natural and textural images based on the extraction of image features and a fast active contour segmentation model. We address the problem of textures where neither the gray-level information nor the boundary information is adequate for object extraction. This is often the case of natural images composed of both homogeneous and textured regions. Because these images cannot be in general directly processed by the gray-level information, we propose a new texture descriptor which intrinsically defines the geometry of textures using semi-local image information and tools from differential geometry. Then, we use the popular Kullback-Leibler distance to design an active contour model which distinguishes the background and textures of interest. The existence of a minimizing solution to the proposed segmentation model is proven. Finally, a texture segmentation algorithm based on the Split-Bregrnan method is introduced to extract meaningful objects in a fast way. Promising synthetic and real-world results for gray-scale and color images are presented.展开更多
An adaptive digital image watermark algorithm with strong robustness based on gray-scale morphology is proposed in this paper.The embedded strategies include:The algorithm seeks and extracts adaptively the image stron...An adaptive digital image watermark algorithm with strong robustness based on gray-scale morphology is proposed in this paper.The embedded strategies include:The algorithm seeks and extracts adaptively the image strong texture regions.The algorithm maps the image strong texture region to the wavelet tree structures, and embeds adaptively watermark into the wavelet coefficients corresponding to the image's strong texture regions.According to the visual masking features, the algorithm adjusts adaptively the watermark-embedding intensity.Experimental results show the algorithm is robust to compression, filtering, noise as well as strong shear attacks.The algorithm is blind watermark scheme.The image strong texture region extraction method based on morphology in this algorithm is simple and effective and adaptive to various images.展开更多
文摘根据CSLBP(center-symmetric local binary pattern)和Uniform LBP(local binary pattern)特征描述行人局部纹理互补性的特点,提出将二者级联的组合特征用于行人检测:基于灰度图像的纹理特征(hybrid local binary pattern,HLBP)和基于颜色空间的纹理特征(color based hybrid local binary pattern,CHLBP)。实验结果表明,当FPPW=10–4时,HLBP特征的检测率为93.96%,与Uniform LBP和CSLBP特征相比分别提高3.46%和9.68%,基于颜色空间L′C′C′与HIKSVM分类器结合时的检测率高达98.58%。与传统的纹理特征检测方法相比,该特征提高了行人检测精度,降低了误检率,检测性能得到较大幅度的提升。
基金supported by Swiss National Science Foundation Grant #205320-101621supported by ONR N00014-03-1-0071
文摘In this paper, we present an efficient approach for unsupervised segmentation of natural and textural images based on the extraction of image features and a fast active contour segmentation model. We address the problem of textures where neither the gray-level information nor the boundary information is adequate for object extraction. This is often the case of natural images composed of both homogeneous and textured regions. Because these images cannot be in general directly processed by the gray-level information, we propose a new texture descriptor which intrinsically defines the geometry of textures using semi-local image information and tools from differential geometry. Then, we use the popular Kullback-Leibler distance to design an active contour model which distinguishes the background and textures of interest. The existence of a minimizing solution to the proposed segmentation model is proven. Finally, a texture segmentation algorithm based on the Split-Bregrnan method is introduced to extract meaningful objects in a fast way. Promising synthetic and real-world results for gray-scale and color images are presented.
基金Supported by the Technology Key Project of Shanxi Province (2007K04-13)the Application Development and Research Project of Xi’an (YF07017)
文摘An adaptive digital image watermark algorithm with strong robustness based on gray-scale morphology is proposed in this paper.The embedded strategies include:The algorithm seeks and extracts adaptively the image strong texture regions.The algorithm maps the image strong texture region to the wavelet tree structures, and embeds adaptively watermark into the wavelet coefficients corresponding to the image's strong texture regions.According to the visual masking features, the algorithm adjusts adaptively the watermark-embedding intensity.Experimental results show the algorithm is robust to compression, filtering, noise as well as strong shear attacks.The algorithm is blind watermark scheme.The image strong texture region extraction method based on morphology in this algorithm is simple and effective and adaptive to various images.
