Integrating color and spatial feature for content-based image retrieval

In this paper, we present a novel and efficient scheme for extracting, indexing and retrieving color images. Our motivation was to reduce the space overhead of partition-based approaches taking advantage of the fact that only a relatively low number of distinct values of a particular visual feature is present in most images. To extract color feature and build indices into our image database we take into consideration factors such as human color perception and perceptual range, and the image is partitioned into a set of regions by using a simple classifying scheme. The compact color feature vector and the spatial color histogram, which are extracted from the seqmented image region, are used for representing the color and spatial information in the image. We have also developed the region-based distance measures to compare the similarity of two images. Extensive tests on a large image collection were conducted to demonstrate the effectiveness of the proposed approach.     

Motion streaks provide a spatial code for motion direction.

Although many neurons in the primary visual cortex (V1) of primates are direction selective, they provide ambiguous information about the direction of motion of a stimulus. There is evidence that one of the ways in which the visual system resolves this ambiguity is by computing, from the responses of V1 neurons, velocity components in two or more spatial orientations and then combining these velocity components. Here I consider another potential neural mechanism for determining motion direction. When a localized image feature moves fast enough, it should become smeared in space owing to temporal integration in the visual system, creating a spatial signal-a 'motion streak'-oriented in the direction of the motion. The orientation masking and adaptation experiments reported here show that these spatial signals for motion direction exist in the human visual system for feature speeds above about 1 feature width per 100 ms. Computer simulations show that this psychophysical finding is consistent with the known response properties of V1 neurons, and that these spatial signals, when appropriately processed, are sufficient to determine motion direction in natural images.     

基于IKNOS遥感影像的北京城市公园湿地资源调查

将1m分辨率全色影像几何形状数据与4m彩色影像数据融合，对形成的1m彩色影像进行解译探讨水系植被等的遥感影像特征信息，获取北京市各个公园水面面积，植被面积等最新数据。     

基于小波变换和视觉感知特性的盲水印算法

借鉴基于视觉感知特性的图像压缩思想,提出了一种基于小波变换和视觉感知特性的自适应盲水印算法．首先,将载体图像进行小波变换,利用视觉系统的照度掩蔽特性和纹理掩蔽特性把载体小波系数分为9类．其次,对二值水印图像分别进行3次不同的置乱,得到3个随机水印序列．最后依据载体图像小波系数的分类,将3个随机水印序列以不同强度嵌入到3个方向的小波系数中．该算法在抽取水印图像时不需要原始载体图像和原始水印图像．实验结果表明,该算法对裁剪、JPEG压缩、加噪、图像增强、滤波等处理都有很好的抵抗力。     

Miniature eye movements enhance fine spatial detail

Our eyes are constantly in motion. Even during visual fixation, small eye movements continually jitter the location of gaze. It is known that visual percepts tend to fade when retinal image motion is eliminated in the laboratory. However, it has long been debated whether, during natural viewing, fixational eye movements have functions in addition to preventing the visual scene from fading. In this study, we analysed the influence in humans of fixational eye movements on the discrimination of gratings masked by noise that has a power spectrum similar to that of natural images. Using a new method of retinal image stabilization, we selectively eliminated the motion of the retinal image that normally occurs during the intersaccadic intervals of visual fixation. Here we show that fixational eye movements improve discrimination of high spatial frequency stimuli, but not of low spatial frequency stimuli. This improvement originates from the temporal modulations introduced by fixational eye movements in the visual input to the retina, which emphasize the high spatial frequency harmonics of the stimulus. In a natural visual world dominated by low spatial frequencies, fixational eye movements appear to constitute an effective sampling strategy by which the visual system enhances the processing of spatial detail.     

A New Method of Color Edge Detection Based on Local Structure Analysis

Human＇s real life is within a colorful world. Compared to the gray images, color images contain more information and have better visual effects. In today＇s digital image processing, image segmentation is an important section for computers to ＂understand＂ images and edge detection is always one of the most important methods in the field of image segmentation. Edges in color images are considered as local discontinuities both in color and spatial domains. Despite the intensive study based on integration of single-channel edge detection results, and on vector space analysis, edge detection in color images remains as a challenging issue.     

多源遥感影像数据融合的理论与技术

目的 探讨多源遥感影像数据融合技术,包括融合概念、融合原理、融合算法、融合效果评价。方法 对于不同空间分辨率、时间分辨率和波谱分辨率的遥感图像进行综合分析,提出像素级、特征级、决策级遥感影像融合的方法及用信息论的理论建立融合影像评价系统。结果 信息融合可分为3个层次：像素级、特征级和决策级融合;像素级融合信息损失最小,决策级融合信息损失最大;像素级融合精度最高,决策级融合精度最低。结论 对同一地区的多源遥感影像数据进行融合,可以产生比单一信息源更准确、更完全、更可靠的估计和判断,可以提高影像的空间分辨率和清晰度,有效提高遥感影像数据的利用率等。     

图像分割与平稳小波变换法融合红外与可见光图像

为改善红外与可见光图像融合后的视觉效果,提出一种基于图像分割和平稳小波变换的图像融合方法。首先,结合最大类间方差方法与边缘检测方法,将红外图像分割为背景部分和目标部分;然后利用平稳小波变换对红外图像的背景部分与可见光图像分别进行多尺度分解,低频部分采用区域空间频率取大融合准则,高频部分采用绝对值取大融合准则,对多尺度分解后的各层进行融合,再利用平稳小波逆变换得到融合结果;最后,对该融合结果与红外图像的目标部分采用加权求和的融合准则进行融合,得到最终的红外与可见光融合图像。实验结果表明,通过提出的方法进行红外与可见光图像融合,不仅很好地突出红外图像的目标信息,还较好地体现可见光图像的场景细节信息,视觉效果明显改善;其标准差、信息熵、互信息均优于拉普拉斯金字塔变换和小波变换等传统的融合算法。实验结果验证了方法的有效性。     

基于时空纹理图象能量分析的运动和深度估计

本文提出用于运动理解和景深估计的三维时空纹理图象的运动能量方法，研究时空纹理和物体运动的关系，并从时域采样的角度研究了对应性问题。在此基础上，着重讨论了在已知运动的情况下，由时空运动纹理的频谱分布来定性估计深度的算法，并给出实验结果和分析。对于一般的运动，作者提出了采用图象注视变换和时域重采样的方法，使复杂的三维时空图象的分析简化为二维时空纹理的方向检测。本文研究可广泛应用于视觉导航、自动检测、安全监视和视频图象分类等领域。     

一种空域与小波域相结合的水印算法

提出了一种空域与小波域相结合的水印算法．算法采用有意义的二值灰度图像作为水印，首先将水印分成2部分，一部分包含重要信息，另一部分包含次要信息．然后利用二维离散小波变换对原始载体图像进行多级分解，并对含重要信息的水印分量进行Arnold变换，以达到置乱图形和消除空间相关性韵目的，接着将水印数据叠加到载体图像的小波分解系数上，实现水印的嵌入．对于含次要信息的水印分量则通过LSB算法嵌入到经逆向小波变换生成的图像的空域中．实验证明：该算法对／PEG有损压缩、噪声和剪切等常见水印攻击显示出了较强的鲁棒性．     

Retinal light adaptation--evidence for a feedback mechanism

Light adaptation is the adjustment of retinal response properties to variations in ambient illumination. It enables the encoding of visual information over a millionfold intensity range, from moonlight to broad daylight, despite the relatively small dynamic range of response of visual neurones. We have studied the effects of light adaptation on the dynamics and sensitivity of visual responses of neurones in the turtle retina, by measuring the responses of horizontal cells in the retina to light which was modulated with a sinusoidal time course around various mean levels. As a quantitative measure of the transduction from light to neural signals, we calculated the gain of response at each frequency. Gain is defined as the amplitude of the modulated response component divided by the amplitude of light modulation. We report here that the gain (mV photon-1) at low temporal frequencies decreased as the mean light level increased. Over a 2 log-unit range of mean light levels, low-frequency gain was inversely proportional to the mean light level, as in Weber's law. However, at high temporal frequencies, the gain was almost independent of mean light level. Our results are reminiscent of Kelly's results on human temporal-frequency sensitivity in various states of light adaptation. We found that a family of horizontal-cell temporal frequency responses, measured at various mean light levels, could be accounted for by a negative feedback model in which the feedback strength is proportional to mean light level.     

Perceptual consequences of centre-surround antagonism in visual motion processing

Centre-surround receptive field organization is a ubiquitous property in mammalian visual systems, presumably tailored for extracting image features that are differentially distributed over space. In visual motion, this is evident as antagonistic interactions between centre and surround regions of the receptive fields of many direction-selective neurons in visual cortex. In a series of psychophysical experiments we make the counterintuitive observation that increasing the size of a high-contrast moving pattern renders its direction of motion more difficult to perceive and reduces its effectiveness as an adaptation stimulus. We propose that this is a perceptual correlate of centre-surround antagonism, possibly within a population of neurons in the middle temporal visual area. The spatial antagonism of motion signals observed at high contrast gives way to spatial summation as contrast decreases. Evidently, integration of motion signals over space depends crucially on the visibility of those signals, thereby allowing the visual system to register motion information efficiently and adaptively.     

Dynamic properties of neurons in cortical area MT in alert and anaesthetized macaque monkeys.

In order to see the world with high spatial acuity, an animal must sample the visual image with many detectors that restrict their analyses to extremely small regions of space. The visual cortex must then integrate the information from these localized receptive fields to obtain a more global picture of the surrounding environment. We studied this process in single neurons within the middle temporal visual area (MT) of macaques using stimuli that produced conflicting local and global information about stimulus motion. Neuronal responses in alert animals initially reflected predominantly the ambiguous local motion features, but gradually converged to an unambiguous global representation. When the same animals were anaesthetized, the integration of local motion signals was markedly impaired even though neuronal responses remained vigorous and directional tuning characteristics were intact. Our results suggest that anaesthesia preferentially affects the visual processing responsible for integrating local signals into a global visual representation.     

视觉导航的多尺度全方位时空图象综合理解方法

本文提出基于全方位时空图象的多尺度视觉导航方法。其基本思想是根据具体视觉任务（道路识别、障碍物检测和全局定位）的要求，采用不同时空尺度的视觉传感器和处理方法，将空域大尺度的全方位“环视”，小尺度的双目“注视”和时域大尺度的时空“远视”相结合，综合完成道路图象的理解。设计了适于机器处理的特殊传感器和快速有效的处理方法，并利用图象级的不变性避免了困难的图象分割和三维恢复。     

基于注意力机制和感知损失的遥感图像去噪

遥感图像去噪对于遥感图像在后续的分类、检测等任务中有着非常重要的作用.为了让去噪后的图像更好地保留边缘细节信息,同时增强深度网络对噪声污染区域的辨识能力,本文结合注意力机制以及感知损失来处理遥感图像,提出了一种新的基于残差自编码器的遥感图像去噪网络ARED-VGG.考虑到遥感图像中不同地物大小不同,该网络首先同时使用图像的空间和光谱信息来提取多尺度特征;然后使用残差自编码器网络结构来进行图像空间-光谱多尺度图像重建.为了增加网络的辨识能力,更多地关注网络中提取的高频特征,网络引入了注意力机制.同时为了让去噪后的结果更符合视觉感观,使用了感知损失混合均方误差作为损失函数.从实验结果知,本文所提出的方法在噪声去除和纹理细节保留方面与NLM3D、BM4D、LRMR、HSID-CNN和3DADCNN相比表现更优.在Washington DC mall遥感图像数据集上进行了仿真实验,平均峰值信噪比以及平均结构相似性指标都有较好的结果;在AVIRIS Indian Pines数据集上进行了真实数据实验,以去噪后的结果地物分类指标作为验证,整体分类精度以及Kappa系数分别为96.90%和0.9647;对网络结构进行了消融实验,在两个数据集下,本文所提出的网络结构都能获得更优的结果.本文提出基于注意力机制和感知损失的深度神经网络进行遥感图像去噪,提高了网络的辨识能力,实现了良好的去噪性能,并且有效保持了图像的细节信息和光谱信息.     

具有细节补偿和色彩恢复的多尺度Retinex色调映射算法

针对带色彩恢复的多尺度Retinex色调映射算法(MSRCR)在分离图像光照信息时未保留部分细节信息,导致结果图像出现细节模糊和颜色失真的问题,提出了一种具有细节补偿和色彩恢复的多尺度Retinex色调映射算法(MSRCD).该算法利用Retinex理论的基本原理将高动态范围图像分为反射层和光照层,先使用双边滤波从图像光照层中提取出细节信息进行补偿,然后从图像的反射层中分离出基本层信息并进行自适应调整,压缩其动态范围,最后通过色彩校正还原图像颜色.实验结果表明,与MSRCR算法及基于双边滤波的算法相比,MSRCD算法的处理结果保留了更丰富的细节信息,色彩逼近于真实场景且避免了光晕的产生.     

可见光与红外图像彩色融合中的热目标增强方法

研究树林、陆地场景下可见光与红外图像彩色融合中的热目标增强方法.利用拮抗视觉特性在YUV空间融合可见光与红外图像,增强二者的独有特征;在色彩传递阶段,使用3×3大小的十字形窗口定位红外图像的热目标,当窗口的局部均值大于红外图像的整体均值时,确定该像素点处于热目标区,表征红色与亮度(Y)差异的V分量值根据局部均值与整体均值之比进行增强.仿真结果表明,算法用浓烈的红色突出红外图像的热目标,有效地提醒观察者注意感兴趣的目标,同时,可见光图像的场景细节被赋予与白天光照下彩色图像类似的自然色彩,改善了场景的深度感知,提高了形势意识能力.     

新型液晶编码光栅立体视觉传感器

物体形面视觉测量中，为了获取更多且准确的图像对应信息，特别是立体像对的匹配信息，需要通过视觉传感器对测量图像进行调制．该文介绍了研究设计的新型液晶编码光栅形面测量视觉传感器，采用东福EDMl288液晶模块和光学投影照明系统，并且通过计算机通讯控制液晶编码，无机械运动，测量速度快，可得到大量三维测量点信息．实验验证，这种传感器可靠性高，可操作性强，且具有较高测量精度，三维测量点拟合平面的平均误差为0．103mm，可以实现物体的形面测量．     

人眼动态视觉特性分析

人眼作为电视系统的接收终端,其视沉特性直接决定着对电视系统的特性要求;也是对电视图象信号进行频带压缩处理的基础,本文基于人眼的视觉特性,提出了人眼视觉响应模型,分析了人眼动态视觉特性,分析表明,由于视觉残留的影响,人眼对活动图象的空间分辨力将降低,而对时变图象则表现为对层次的分辨力下降,其结果都降低了对单幅图象的有效息信量,从而为对视频图象信号的频带压缩提供了可能。     

基于跳跃马尔可夫模型的红外点目标检测方法

提出一种基于跳跃马尔可夫模型的红外点目标检测方法,根据模型设计了采用粒子滤器的模型参数与状态估计算法.该方法既能用于图像的空域滤波,也能用于序列图像的时域滤波.实验表明采用该方法进行空域/时域滤波可以极大地抑制杂波背景并增强点目标的可检测性.