提出了一种新的基于系数状态表的SPIHT(LPS-SPIHT,list of p ixel stata-set partition ing in h ierarch icaltrees)图像压缩编码算法,该算法具有以下5个特点:第一,定义了一种扩展的空间方向树,使1个结点含有2×2相邻的4个系数,并...提出了一种新的基于系数状态表的SPIHT(LPS-SPIHT,list of p ixel stata-set partition ing in h ierarch icaltrees)图像压缩编码算法,该算法具有以下5个特点:第一,定义了一种扩展的空间方向树,使1个结点含有2×2相邻的4个系数,并将基本EZW(嵌入式小波零树)的符号定义应用于扩展树;第二,用1个廉价的系数状态表代替了SPIHT算法中的LIS(不重要集合表)、LIP(不重要像素表)、LSP(重要像素表)等3个数据表,节省了内存;第三,通过扫描系数状态表,可一次性完成对图像数据的编码,使分类过程与细化过程合二而一;第四,利用一种树指数避免了重复计算,提高了处理速度;第五,通过重新组织编码过程,省去了对大量可推知位的编码,提高了压缩效率。实践证明,与目前公认的最为有效的SPIHT算法相比,该算法不仅性能优越,而且计算简单,容易实现。展开更多
We present a robust connected-component (CC) based method for automatic detection and segmentation of text in real-scene images. This technique can be applied in robot vision, sign recognition, meeting processing and ...We present a robust connected-component (CC) based method for automatic detection and segmentation of text in real-scene images. This technique can be applied in robot vision, sign recognition, meeting processing and video indexing. First, a Non-Linear Niblack method (NLNiblack) is proposed to decompose the image into candidate CCs. Then, all these CCs are fed into a cascade of classifiers trained by Adaboost algorithm. Each classifier in the cascade responds to one feature of the CC. Proposed here are 12 novel features which are insensitive to noise, scale, text orientation and text language. The classifier cascade allows non-text CCs of the image to be rapidly discarded while more computation is spent on promising text-like CCs. The CCs passing through the cascade are considered as text components and are used to form the segmentation result. A prototype system was built, with experimental results proving the effectiveness and efficiency of the proposed method.展开更多
A hierarchical non-linear method for image registration was presented, which integrates image segmentation and registration under a variational framework. An improved deformable model is used to simultaneously segment...A hierarchical non-linear method for image registration was presented, which integrates image segmentation and registration under a variational framework. An improved deformable model is used to simultaneously segment and register feature from multiple images. The objects in the image pair are segmented by evolving a single contour and meanwhile the parameters of affine registration transformation are found out. After that, a contour-constrained elastic registration is applied to register the images correctly. The experimental results indicate that the proposed approach is effective to segment and register medical images.展开更多
文摘提出了一种新的基于系数状态表的SPIHT(LPS-SPIHT,list of p ixel stata-set partition ing in h ierarch icaltrees)图像压缩编码算法,该算法具有以下5个特点:第一,定义了一种扩展的空间方向树,使1个结点含有2×2相邻的4个系数,并将基本EZW(嵌入式小波零树)的符号定义应用于扩展树;第二,用1个廉价的系数状态表代替了SPIHT算法中的LIS(不重要集合表)、LIP(不重要像素表)、LSP(重要像素表)等3个数据表,节省了内存;第三,通过扫描系数状态表,可一次性完成对图像数据的编码,使分类过程与细化过程合二而一;第四,利用一种树指数避免了重复计算,提高了处理速度;第五,通过重新组织编码过程,省去了对大量可推知位的编码,提高了压缩效率。实践证明,与目前公认的最为有效的SPIHT算法相比,该算法不仅性能优越,而且计算简单,容易实现。
文摘We present a robust connected-component (CC) based method for automatic detection and segmentation of text in real-scene images. This technique can be applied in robot vision, sign recognition, meeting processing and video indexing. First, a Non-Linear Niblack method (NLNiblack) is proposed to decompose the image into candidate CCs. Then, all these CCs are fed into a cascade of classifiers trained by Adaboost algorithm. Each classifier in the cascade responds to one feature of the CC. Proposed here are 12 novel features which are insensitive to noise, scale, text orientation and text language. The classifier cascade allows non-text CCs of the image to be rapidly discarded while more computation is spent on promising text-like CCs. The CCs passing through the cascade are considered as text components and are used to form the segmentation result. A prototype system was built, with experimental results proving the effectiveness and efficiency of the proposed method.
基金National Natural Science Foundation ofChina (No.60271033)
文摘A hierarchical non-linear method for image registration was presented, which integrates image segmentation and registration under a variational framework. An improved deformable model is used to simultaneously segment and register feature from multiple images. The objects in the image pair are segmented by evolving a single contour and meanwhile the parameters of affine registration transformation are found out. After that, a contour-constrained elastic registration is applied to register the images correctly. The experimental results indicate that the proposed approach is effective to segment and register medical images.
