深度图像中基于轮廓曲线和局部区域特征的3维物体识别

为了提高3维物体目标识别系统的性能及降低计算复杂度,提出一种由粗到细的识别方法。该方法利用深度图像所提供的信息,分两步完成识别过程。首先基于轮廓曲线计算其特征点,并映射到原有轮廓空间,以标志点序列表征原由轮廓进行匹配,在识别初期迅速排除模型库中不相似目标和差异较大的姿态,生成目标候选列表用于精确匹配,以提高识别效率。精确匹配采用一种基于局部区域特征的识别方法,以投票的策略获取最佳结果。局部区域由SIFT算子确定位置和数量,区域特征主要由表面指数和法向量夹角组成,具有平移和旋转不变性。为了更进一步提高效率和降低存储空间,模型库的数据分为轮廓和表面信息两部分,分别以标志位序列和哈希表的形式存储。实验结果表明,该方法具有良好的实时性和识别率,对遮挡和干扰有一定的适应性。     

A novel natural image noise level estimation based on flat patches and local statistics

Multimedia Tools and Applications - This paper proposes a high-precision algorithm for noise level estimation. Different from existing algorithms, we present a new noise level estimation algorithm...     

On the behavior of artificial neural network classifiers inhigh-dimensional spaces

It is widely believed in the pattern recognition field that when a fixed number of training samples is used to design a classifier, the generalization error of the classifier tends to increase as the number of features gets larger. In this paper, we discuss the generalization error of the artificial neural network (ANN) classifiers in high-dimensional spaces, under a practical condition that the ratio of the training sample size to the dimensionality is small. Experimental results show that the generalization error of ANN classifiers seems much less sensitive to the feature size than 1-NN, Parzen and quadratic classifiers     

On Wachspress quadrilateral patches

Wachspress initiated the study of rational basis functions for finite element construction over quadrilaterals and more general polygonal and curved elements. Later Apprato et al. (1979) and Gout (1979, 1985) studied the interpolatory and convergence properties of lower degree rational finite elements and their applications in solving second order boundary value problems. In the present paper we introduce higher degree Wachspress functions by an iterative technique and study their properties from the point of view of applications to surface fitting problems. It is indeed remarkable to note that these functions possess properties similar to tensor product Bernstein polynomials and hence could be effectively used to generate quadrilateral patches.     

An adaptive dynamic range compression with local contrast enhancement algorithm for real-time color image enhancement

Journal of Real-Time Image Processing - Dynamic range compression has become an important function used in modern digital video cameras to improve visual quality of color images suffered from low...     

A local stereo matching algorithm based on weighted guided image filtering for improving the generation of depth range images

Stereo matching is a challenging problem and highly accurate depth image is important in different applications. The main problem is to estimate the correspondence between two pixels in a stereo pair. To solve this problem, in the last decade, several cost aggregation methods aimed at improving the quality of stereo matching algorithms have been introduced. We propose a new cost aggregation method based on weighted guided image filtering (WGIF) for local stereo matching. The proposed algorithm solves multi-label problems in three steps. First, the cost volume is constructed using pixel-wise matching cost computation functions. Then, each slice of the cost volume is independently filtered using the WGIF, which substitutes for the smoothness term in the energy function. Finally, the disparity of any pixel is simply computed. The WGIF uses local weights based on a variance window of pixels in a guidance image for cost volume filtering. Experimental results using Middlebury stereo benchmark verify that the proposed method is effective due to a high quality cost volume filter.     

Microscopic image restoration based on tensor factorization of rotated patches

In microscopic image processing for analyzing biological objects, structural characters of objects such as symmetry and orientation can be used as a prior knowledge to improve the results. In this study, we incorporated filamentous local structures of neurons into a statistical model of image patches and then devised an image processing method based on tensor factorization with image patch rotation. Tensor factorization enabled us to incorporate correlation structure between neighboring pixels, and patch rotation helped us obtain image bases that well reproduce filamentous structures of neurons. We applied the proposed model to a microscopic image and found significant improvement in image restoration performance over existing methods, even with smaller number of bases.     

宽动态范围图像快速增强算法

提出了一种针对宽动态范围图像的快速增强算法。该算法基于反转后宽动态范围图像与雾天图像有相似特性的结论,对宽动态范围图像进行反转处理,将优化后基于暗原色先验理论的去雾算法用于反转后的图像。通过改进透射率估计方法和简化大气光估计,显著地降低了算法的复杂度。实验结果表明,该算法有效地改善了宽动态范围图像的视觉效果,并且算法的处理速度很快,能够满足实时应用的要求。     

Functional behavior in data spaces

Summary In order to verify that a nondeterministic sequential program is partially correct it is sufficient to establish the conjunction of two constituent properties: weak partial correctness and functional, that is reproducible, behavior. It is possible to continue this divide-and-conquer strategy for the concept of functional behavior. If the nondeterministic sequential program is derived from a set of interacting parallel processes then the functional behavior of the former can be expressed in terms of two weaker complementary properties of the latter: weak functional behavior and input/output liveness. The only remaining issue is input/output dependability: the absence of input/output livelock. The theoretical framework of data spaces is used to derive closure theorems for these constituent properties. For instance, it is shown that a system of weakly functional processes is again weakly functional.This research is supported in part by the Office of Naval Research under Contract No. N00014-77-C-0536 through the University of Southern CaliforniaA preliminary version of the paper was presented at the Symposium on Formal Methods and Mathematical Tools for Software Design, Mathematical Research Institute, Oberwolfach, West Germany, Nov. 21–27, 1976     

基于多字典学习和图像块映射的超分辨率重建

针对单一冗余字典在稀疏表示图像超分辨率重建结果出现不清晰、伪影以及重建过程编码效率不高、运算时间过长的问题,提出一种基于多字典学习和图像块映射的超分辨率重建方法。该方法在传统稀疏表示的框架下,首先探索局部图像块的梯度结构信息,按梯度角度将训练样本块分类;然后为每个子类样本集学习高低分辨率字典对,再结合最近邻思想应用生成的字典,为每个子类计算从低分辨率块到高分辨率块映射的函数;最后将重建过程简化为输入块和映射函数的乘积,在保证提高重建质量的同时减少了图像重建的时间。实验结果表明,所提算法在视觉效果有较大的提升,同时与锚点邻域回归算法相比,评价参数峰值信噪比(PSNR)平均提高约0.4 dB。     

The importance of native panel contrast and local dimming density on perceived image quality of high dynamic range displays

We evaluated the perceived image quality of High Dynamic Range (HDR) content rendered using different types of local dimming and organic light‐emitting diode (OLED) displays. Using an OLED display that is capable of achieving high contrast at a pixel level, we emulated local dimming displays to evaluate their image quality. In a set of subjective experiments, observers compared HDR images and videos rendered with different local dimming densities and native panel contrast. There was a strong effect of panel contrast on perceived quality and also a strong trend toward preference for a larger number of dimming zones. We also evaluated the panel contrast and number of local dimming zones necessary to achieve image quality comparable with OLED. The findings of these experiments demonstrated that the use of a high‐contrast panel remains of critical importance. Also, the preference for panel rendering mode remains robust to normal levels of indoor ambient light.     

Shape analysis of local facial patches for 3D facial expression recognition

In this paper we address the problem of 3D facial expression recognition. We propose a local geometric shape analysis of facial surfaces coupled with machine learning techniques for expression classification. A computation of the length of the geodesic path between corresponding patches, using a Riemannian framework, in a shape space provides a quantitative information about their similarities. These measures are then used as inputs to several classification methods. The experimental results demonstrate the effectiveness of the proposed approach. Using multiboosting and support vector machines (SVM) classifiers, we achieved 98.81% and 97.75% recognition average rates, respectively, for recognition of the six prototypical facial expressions on BU-3DFE database. A comparative study using the same experimental setting shows that the suggested approach outperforms previous work.