Mapping vision algorithms to parallel architectures

Some of the problems encountered in mapping a parallel algorithm are examined, emphasizing mappings of vision algorithms onto mesh, hypercube, mesh-of-trees, pyramid, and parallel random-access machines (PRAMs) having many simple processors, each with a small amount of memory. Approaches that have been suggested include simulating the ideal architectures, and using general data movement operations. Each of these is shown to occasionally produce unacceptably inefficient implementations. It appears that as long as PRAMs cannot achieve the desired cost and performance goals, programmers must contend with carefully designing algorithms for specific architectures     

Three vision algorithms for acquiring workpieces from bins

Automation requires workpieces to be well organized. Current machine feeding technology is dominated by the use of human labor, mechanical feeders, or orientation preservation. All these techniques have problems and limitations that make an alternate solution desirable. Three vision algorithms are presented which enable a computer-controlled robot to acquire a single piece from a bin containing randomly placed identical pieces. The algorithms do not try to solve the problem by identifying the position and orientation of a piece in the bin. Rather, the algorithms recognize where there is a section of any piece which a particular type of gripper would be able to grasp with a reasonable chance of success. Thus the algorithms are holdsite driven and depend on the piece and gripper type. Vacuum cup and parallel-jaw gripper types are treated. The heuristically defined holdsites for vacuum cup grippers are patches of smooth surfaces; for parallel-jaw grippers, holdsites are opposing, linear, or curvilinear parallel edges. The three-dimensional position of holdsites cannot be computed from a single image of the bin. Thus once a hold-sight is identified, the gripper is made to travel along a line-of-sight path. The missing degree of freedom, distance from the holdsite to the camera, is obtained by a contact or noncontact proximity sensor in the gripper. Acquisition requires two other sensors: a grasping sensor to indicate success; and a mechanical overload sensor, to abort the attempt in case of error.     

彩色夜视成像处理算法的新进展

彩色夜视成像技术是当前国内外重点发展的关键技术之一,介绍了彩色夜视成像处理算法的分类、特点及其研究进展,分析了国外一些彩色夜视系统,并重点介绍了笔者在自然感彩色夜视算法及自然感彩色夜视成像系统方面的研究进展。     

基于计算机视觉的无人机目标检测算法综述

随着人工智能的发展浪潮,计算机视觉得到了飞速发展,利用基于深度学习的卷积神经网络可以较为快速的实现目标检测.本文基于无人机领域目标检测现状的难点,通过对目标检测相关算法研究发展进行概优缺点分析,最后阐述了近期无人机对已存在问题的相关前沿研究.综合来看,目标检测相关算法对于自然图像的处理完全契合了无人机发展的基本要求.     

无参考图像质量评价

图像质量评价是近几年图像处理领域比较热门的研究课题。目前,许多学者已经提出了各种各样的无参考质量评价方法。对无参考方法进行综述,详细介绍BIQI,BLIINDS-Ⅱ,BRISQUE,DESIQUE,DIIVINE,NIQE,SSEQ等无参考质量评价方法,并在LIVE和TID2008数据库上进行实验分析,最后根据分析的结果探讨图像质量评价的发展方向。     

基于视觉的运动损伤评估方法

传统损伤评估中使用的是医学临床数据作为评估标准,由于临床医学数据需要对损伤的部位进行处理后才能得出结论,故传统损伤的评估过程存在极大的延时性。为此提出基于视觉的运动损伤评估方法,使用视觉数据代替医学临床数据,修订了评估标准降低传统评估过程延时性。引入视觉s EMG技术对损伤部位进行损伤数据采集,对视觉数据实施转化分析完成运动损伤的评估。实验结果表明,改进评估方法在评估过程中没有延时性,评估精准度要优于传统评估方法,视觉数据能够记录损伤过程对其他运动员的类拟损伤可以提供康复建议。     

Dual-branch vision transformer for blind image quality assessment

Blind image quality assessment (BIQA) has always been a challenging problem due to the absence of reference images. In this paper, we propose a novel dual-branch vision transformer for BIQA, which simultaneously considers both local distortions and global semantic information. It first extracts dual-scale features from the backbone network, and then each scale feature is fed into one of the transformer encoder branches as a local feature embedding to consider the scale-variant local distortions. Each transformer branch obtains the context of global image distortion as well as the local distortion by adopting content-aware embedding. Finally, the outputs of the dual-branch vision transformer are combined by using multiple feed-forward blocks to predict the image quality scores effectively. Experimental results demonstrate that the proposed BIQA method outperforms the conventional methods on the six public BIQA datasets.     

夜视融合图像质量主观评价方法

为建立通用、客观的融合图像质量评价方法,在分析图像质量主观评价方法基础上,研究了融合图像质量主观评价对象、评价条件、评价指标和数据处理等关键环节,结合目标探测识别与场景理解两个典型的视觉任务,提出了目标可探测性和细节可分辨性两个主观评价指标,并研究了图像整体感知质量与这两个指标的相关性。对3种不同场景的微光与红外融合图像,采用9种融合方法获得的189幅融合图像的统计分析结果显示,细节可分辨性和图像整体感知质量相关性好,而目标可探测性虽然和图像整体感知质量相关性较差,但在基于具体视觉任务的融合图像质量评价过程中,仍可作为有效评价指标之一。     

夜视融合图像质量客观评价方法

为建立通用、客观的融合图像质量评价方法，在分析图像质量评价与融合图像质量评价关系基础上，给出了图像质量评价与融合图像质量评价的一般表达式。依据信息理论和结构相似度评价方法，对建立的4种客观评价指标，采用4种融合方法获得的36幅融合图像进行了主观评价实验，统计分析结果显示，结合人类视觉系统的客观评价方法优于熵、交互信息量等评价指标，但仍未达到高度的主客观一致性，说明构建通用、高效、主客观一致性好的融合图像质量评价指标存在较大难度，同时对可能存在的原因进行了分析。     

动作识别算法的评估策略探讨

以时空兴趣点特征和支持向量机(SVM)分类器识别方法为基本算法,在广泛使用的公开动作数据集KTH上,从不同角度考察评估策略对动作识别算法性能的影响。实验表明,当采用不同的交叉实验方法时,算法性能的波动最大达到10.5%,而不同数据集划分方法对算法性能的影响则达到11.87%。因此,通过量化分析得出的结论,可以充分地比较现有算法的真实差异,并为设计合理的评估策略提供参考。     

Convergence and stability assessment of Newton-Kantorovichreconstruction algorithms for microwave tomography

For newly developed iterative Newton-Kantorovitch reconstruction techniques, the quality of the final image depends on both experimental and model noise. Experimental noise is inherent to any experimental acquisition scheme, while model noise refers to the accuracy of the numerical model, used in the reconstruction process, to reproduce the experimental setup. This paper provides a systematic assessment of the major sources of experimental and model noise on the quality of the final image. This assessment is conducted from experimental data obtained with a microwave circular scanner operating at 2.33 GHz. Targets to be imaged include realistic biological structures, such as a human forearm, as well as calibrated samples for the sake of accuracy evaluation. The results provide a quantitative estimation of the effect of experimental factors, such as temperature of the immersion medium, frequency, signal-to-noise ratio, and various numerical parameters     

融合机器视觉与惯性传感器的智能康复评估技术

为缓解患有运动功能障碍的患者不方便频繁前往医院进行康复评估的问题,本文提出一种基于"互联网+"的居家康复评估技术,为康复医师提供参考,节约康复医师和患者的时间,缓解医疗资源不足.针对脑卒中患者上肢康复量表的评估问题,仅使用惯性传感器或者机器视觉建立的长短期记忆人工神经网络分类模型在测试样本中准确率分别为55.6％、92...     

Growth of the Greenland ice sheet: a performance assessment ofaltimeter retracking algorithms

The authors compare the performance of different altimeter retracking algorithms for measuring ice sheet elevations and growth rates. The results show that the threshold, ESA, and S/V retracking algorithms produce growth rates that are 30-55% smaller than those produced by the NASA algorithm. Based upon a comparison of crossover-point standard deviations, the analysis indicates that the surface elevation estimates produced by these algorithms are more repeatable than the NASA surface elevations. An analysis of the NASA algorithm shows that a mixing of its 5 and 9 parameter functional fits on the crossover-point altimeter waveforms occurs in over 70% of the crossover data. The mixing of the functional fits is shown to reduce the repeatability of the NASA elevations and this may be responsible for the larger estimates of ice sheet growth produced by the NASA retracking algorithm. The extremely close agreement between the standard deviations and the growth-rate estimates from the threshold, ESA, and S/V retracking algorithms: lead the authors to conclude that 0.10 m/yr is a more accurate estimate of the growth of the Greenland ice sheet from 1978-1987 (south of 72°N)     

A technical and economic assessment of computer vision for industrial inspection and robotic assembly

The use of computer vision to detect, measure, and perhaps guide the assembly of man-made components is potentially a very significant research and development area. The efficacy of these techniques for any given application depends on both technical and economic considerations. This paper will explore both these considerations using appropriate generic examples. It is our goal to first present a concise discussion of the present state of many technical and economic factors and then extrapolate these factors into the future for the purpose of guiding further investigations.     

基于生物视觉标准模型特征的无参考型图像质量评价方法

鉴于生物视觉特征对于图像的良好表征能力,提出了一种基于生物视觉特征的无参考型图像质量评价方法。对生物视觉ST模型进行了研究和分析,完成了对图像的稀疏化表示;利用最小二乘支持向量机回归方法训练生物视觉特征到图像质量的映射关系,获得能够预测图像质量的回归器;通过学习的回归器完成了对图像质量的评价。基于LIVE图像库的实验结果表明,该方法对于特定失真和交叉失真的预测误差分别为2%和5%左右,并且与目前技术条件下的质量评价方法相比具有很好的精确性和单调性。     

基于社区搜索模块排序算法的网页设计教学质量评估系统

教学质量的评估可推动教学机构教学工作全面改进与提升,促进教学改革纵深发展。而针对以往教学质量评估大多通过学生单方面反馈来判定,评估结果较为片面,且评估精度不高。将网页设计作为评估领域,设计一种基于社区搜索模块排序算法网页设计教学质量评估系统,通过数据采集模块获取有关网页主题、网页栏目规划、网页内容、网页版面设计及网页技术的数据,传输至社区搜索模块;社区搜索模块将评估数据分为教师社区评估数据、学生社区评估数据及管理员社区评估数据,从多个角度分析教学质量,并采用多因素排序算法完成网页设计教学质量评估。实验结果表明,所设计系统可有效评估多个高校网页设计教学质量,且评估精度高达99.89%。     

Performance assessment of feature detection algorithms: a methodology and case study on corner detectors

We describe a generic methodology for evaluating the labeling performance of feature detectors. We describe a method for generating a test set and apply the methodology to the performance assessment of the three well-known corner detectors of L. Kitchen and A. Rosenfeld (1982), of K. Paler et al. (1984), and of C. Harris and M. Stephens (1988). The labeling deficiencies of each of these detectors is related to their discrimination ability between corners and various of the features which comprise the class of noncorners.     

Performance dependency of retinal image quality assessment algorithms on image resolution: analyses and solutions

Retinal image quality assessment (RIQA) is the first step performed in retinal image processing systems necessary to assure that the processed images are suitable for analysis and medical diagnosis. RIQA algorithms created for controlled environments can result in degraded performance for cross-dataset experiments in which the train and test images have different resolutions. The effect of image resolution on the performance of four different RIQA algorithms, chosen to include generic, segmentation, and transform-based quality features, is studied using datasets of various resolutions. Analyses showed that for cross-dataset classifications, the performance of some RIQA algorithms was reduced by up to 50% in cases where the train and test dataset image resolutions were significantly different. A statistical analysis was conducted to study how the retinal image quality features are affected by image resolution which resulted in their categorization into resolution-dependent and resolution-independent features. Feature scaling was then introduced to overcome the transform-based RIQA algorithm’s cross-dataset performance degradation resulting in a 100% performance enhancement. Based on this study, the investigation and enhancement of the cross-dataset performance of RIQA algorithms are recommended as a standard part of their design in order to assure their performance reliability in processing images of various resolutions.     

Raindrop axis ratios and size distributions in Florida rainshafts:an assessment of multiparameter radar algorithms

Eleven penetrations of rainshafts by the University of Wyoming King Air (WKA) aircraft equipped with a two-dimensional (2D) optical array probe are studied in coordination with multiparameter radar measurements from the National Center for Atmospheric Research (NCAR) CP-2 radar collected in a multicellular storm that occurred on August 8, 1991, of the Convective and Precipitation/Electrification (CaPE) experiment. A comparison is made between the mass-weighted mean diameter (D_m) and rainrate (R) computed from the nine-size spectra and their estimates from multiparameter radar algorithms based on Z_dr , and Z_h. It was found that D_m could be estimated with a mean bias of 0.07 mm and a standard deviation of 0.35 mm. Rainrates (in the range of 10-60 mmh^-1) could be estimated from Z_h, and Z_dr with a mean bias of 1-4% and fractional standard error (FSE) of 30-40% depending on the estimator used. Raindrop axis ratios are analyzed as a function of volume equivalent spherical diameter (D_eq) in the range 2-6 mm. The mean axis ratio versus the D_eq relationship was found to be consistent with previous data from the High Plains (from Colorado and Montana). A study of fluctuations of axis ratio (about their mean value) showed that most drops have axis ratios close to their mean values with oscillation amplitudes to be typically ±10% in axis ratio, again consistent with the earlier High Plains results     

Optimization of image coding algorithms and architectures usinggenetic algorithms

This paper addresses the application of genetic algorithm (GA)-based optimization techniques to problems in image and video coding, demonstrating the success of GAs when used to solve real design problems with both performance and implementation constraints. Issues considered include problem representation, problem complexity, and fitness evaluation methods. For offline problems, such as the design of two-dimensional filters and filter banks, GAs are shown to be capable of producing results superior to conventional approaches. In the case of problems with real-time constraints, such as motion estimation, fractal search and vector quantization codebook design, GAs can provide solutions superior to those reported using conventional techniques with comparable implementation complexity. The use of GAs to jointly optimize algorithm performance in the context of a selected implementation strategy is emphasized throughout and several design examples are included