Unsupervised real-time constrained linear discriminant analysis to hyperspectral image classification

We have proposed a constrained linear discriminant analysis (CLDA) approach for classifying the remotely sensed hyperspectral images. Its basic idea is to design an optimal linear transformation operator which can maximize the ratio of inter-class to intra-class distance while satisfying the constraint that the different class centers after transformation are aligned along different directions. Its major advantage over the traditional Fisher's linear discriminant analysis is that the classification can be achieved simultaneously with the transformation. The CLDA is a supervised approach, i.e., the class spectral signatures need to be known a priori. But, in practice, these informations may be difficult or even impossible to obtain. So in this paper we will extend the CLDA algorithm into an unsupervised version, where the class spectral signatures are to be directly generated from an unknown image scene. Computer simulation is used to evaluate how well the algorithm performs in terms of finding the pure signatures. We will also discuss how to implement the unsupervised CLDA algorithm in real-time for resolving the critical situations when the immediate data analysis results are required.     

Implementation of real-time constrained linear discriminant analysis to remote sensing image classification

In this paper, we investigate the practical implementation issues of the real-time constrained linear discriminant analysis (CLDA) approach for remotely sensed image classification. Specifically, two issues are to be resolved: (1) what is the best implementation scheme that yields lowest chip design complexity with comparable classification performance, and (2) how to extend CLDA algorithm for multispectral image classification. Two limitations about data dimensionality have to be relaxed. One is in real-time hyperspectral image classification, where the number of linearly independent pixels received for classification must be larger than the data dimensionality (i.e., the number of spectral bands) in order to generate a non-singular sample correlation matrix R for the classifier, and relaxing this limitation can help to resolve the aforementioned first issue. The other is in multispectral image classification, where the number of classes to be classified cannot be greater than the data dimensionality, and relaxing this limitation can help to resolve the aforementioned second issue. The former can be solved by introducing a pseudo inverse initiate of sample correlation matrix for R^-1 adaptation, and the latter is taken care of by expanding the data dimensionality via the operation of band multiplication. Experiments on classification performance using these modifications are conducted to demonstrate their feasibility. All these investigations lead to a detailed ASIC chip design scheme for the real-time CLDA algorithm suitable to both hyperspectral and multispectral images. The proposed techniques to resolving these two dimensionality limitations are instructive to the real-time implementation of several popular detection and classification approaches in remote sensing image exploitation.     

Incremental complete LDA for face recognition

The complete linear discriminant analysis (CLDA) algorithm has been proven to be an effective tool for face recognition. The CLDA method can make full use of the discriminant information of the training samples. However, the original implementation of CLDA may not suitable for incremental learning problem. In this paper, we first propose a new implementation of CLDA, which is theoretically equivalent to the original implementation of CLDA but is more efficient than the original one. Then, based on our proposed novel implementation of CLDA, we propose the incremental CLDA method which can accurately update the discriminant vectors of CLDA when new samples are inserted into the training set. Experiments on ORL, AR and PIE face databases show the efficiency of our proposed CLDA algorithms over the original implementation of CLDA.     

Fast real-time onboard processing of hyperspectral imagery for detection and classification

Remotely sensed hyperspectral imagery has many important applications since its high-spectral resolution enables more accurate object detection and classification. To support immediate decision-making in critical circumstances, real-time onboard implementation is greatly desired. This paper investigates real-time implementation of several popular detection and classification algorithms for image data with different formats. An effective approach to speeding up real-time implementation is proposed by using a small portion of pixels in the evaluation of data statistics. An empirical rule of an appropriate percentage of pixels to be used is investigated, which results in reduced computational complexity and simplified hardware implementation. An overall system architecture is also provided.

图像欧氏距离在人脸识别中的应用研究

图像欧氏距离可以嵌入到许多传统的图像分类识别算法中,该嵌入是通过对原始图像的线性变换来实现的,给出了一种基于数据场的图像线性变换方法,将其应用到图像欧氏距离中.实验结果表明,基于数据场的线性变换方法是一种可行的图像线性变换方法,该方法可以完成大尺度图像的线性变换,方便地将图像欧氏距离嵌入到传统人脸识别算法中.     

A VLSI Image Processing Architecture Dedicated to Real-Time Quality Control Analysis in an Industrial Plant

In this paper, we present a VLSI architecture for real-time image processing in quality control industrial applications: automation of the visual inspection phase of mechanical parts treated by the Fluorescent Magnetic Particle Inspection method for structural-defect detection. The VLSI architecture implements a highly constrained neural network tailored for this specific application: the multi-layer perceptron with strictly local connections. The learning of the weights is performed off line by using the adaptive simulated-annealing algorithm. The neural network has been trained on real plant data: recognition results of the training and classification tasks compare favorably with those obtained by expert human operators.The VLSI architecture receives as input the image (taken on-line on the plant) of a mechanical part and it will find out if at least one structural surface defect is present. The VLSI architecture was optimized, through a set of transformations on the high-level VHDL specifications of the neural network algorithm, to reach real-time operating conditions. Following the proposed approach and the designed architecture, we designed and successfully tested a custom VLSI chip for the real-time implementation of the recognition task.     

结合四元数与最小核值相似区的边缘检测

目的 针对传统彩色图像边缘检测方法中未充分利用图像色度信息、颜色模型间非线性转换过程中时间和空间的大量耗费、算法实现复杂等问题,将四元数引入最小核值相似区（SUSAN）算法中,提出一种RGB空间下的结合四元数与最小核值相似区的边缘检测算法。方法 该算法首先对彩色图像进行四元数描述,然后用改进的SUSAN算子进行边缘检测。针对其中单一几何阈值g的限制,以及检测出的边缘较粗等问题,本文采用Otsu算法自适应获取双几何阈值,再对弱边缘点集进行边缘生长,最后根据USAN重心及其对称最长轴来确定边缘局部方向,实现对边缘点的局部非极大值抑制,得到最终细化后的边缘图像。结果 实验选取1幅合成彩色图像及3幅标准图像库图像,与彩色Canny算法、SUSAN算法,及采用单阈值的本文算法进行对比,并采用Pratt品质因数衡量边缘定位精度。本文算法能够检测出亮度相近的不同颜色区域之间的边缘,且提取的边缘比较连续、细致,漏检边缘较少。与公认边缘检测效果较好的彩色Canny算法相比,本文算法的品质因数提高了0.012 0,耗时缩短了2.527 9 s。结论 本文提出了一种结合四元数与最小核值相似区的边缘检测算法,实现了四元数与SUSAN算子的有效融合。实验结果表明,该算法能够提高边缘定位精度,对弱噪声具有较好的抑制能力,适用于对实时性要求不高的低层次彩色图像处理。     

结合区域协方差分析的图像显著性检测

目的 图像显著性检测的目的是为了获得高质量的能够反映图像不同区域显著性程度的显著图,利用图像显著图可以快速有效地处理图像中的视觉显著区域。图像的区域协方差分析将图像块的多维特征信息表述为一个协方差矩阵,并用协方差距离来度量两个图像块特征信息的差异大小。结合区域协方差分析,提出一种新的图像显著性检测方法。方法 该方法首先将输入的图像进行超像素分割预处理;然后基于像素块的区域协方差距离计算像素块的显著度;最后对像素块进行上采样用以计算图像像素点的显著度。结果 利用本文显著性检测方法对THUS10000数据集上随机选取的200幅图像进行了显著性检测并与4种不同方法进行了对比,本文方法估计得到的显著性检测结果更接近人工标定效果,尤其是对具有复杂背景的图像以及前背景颜色接近的图像均能达到较好的检测效果。结论 本文方法将图像像素点信息和像素块信息相结合,避免了单个噪声像素点引起图像显著性检测的不准确性,提高了检测精确度;同时,利用协方差矩阵来表示图像特征信息,避免了特征点的数量、顺序、光照等对显著性检测的影响。该方法可以很好地应用到显著目标提取和图像分割应用中。     

肤色检测技术综述

肤色检测在人脸和手势识别与跟踪、Web图像内容过滤、数据库或因特网中的人物检索和医疗诊断等方面有广泛应用,文中通过分别介绍基于统计和基于物理的两类肤色检测技术,较全面地综述了肤色检测技术,其中对颜色空间选择、静、动态肤色建模方法、肤色反射模型和肤色波谱特性等肤色检测重要环节做了分析,明确了选择颜色空间与特征提取和分类方法的联系,强调了研究肤色波谱特征对基于物理的肤色检测技术的重要性,最后探讨了肤色检测的技术难题和发展趋势。     

Efficiently extracting and classifying objects for analyzing color documents

Conventional objects extraction method are not efficient for color document image with large graphics. For example, the projection profile and connected component based methods scanning the large graphics many times. To display the large graphics are extracted, conventional methods use rectangle to represent it. Thus, scanning into the large graphics is time-consuming. In this paper, a novel system for efficiently analyzing color documents is proposed to solve above mentioned problem. The proposed system includes color transformation, background color determination, objects extraction by top-down method, and objects classification without parameters. The proposed color document analysis system is efficient because it scans only background pixels such that the temporal complexity is O (NB), where NB is the total number of background color pixels. Results of this study demonstrate that this system is more effective and efficient than other methods. Moreover, the proposed algorithm can be run in an embedded environment (such as a mobile device) and processed in real-time system due to its simplicity and efficiency.     

Hybrid fusion and interpolation algorithm with near-infrared image

Silicon-based digital cameras can record visible and near-infrared (NIR) information, in which the full color visible image (RGB) must be restored from color filter array (CFA) interpolation. In this paper, we propose a unified framework for CFA interpolation and visible/NIR image combination. To obtain a high quality color image, the traditional color interpolation from raw CFA data is improved at each pixel, which is constrained by the corresponding monochromatic NIR image in gradient difference. The experiments indicate the effectiveness of this hybrid scheme to acquire joint color and NIR information in real-time, and show that this hybrid process can generate a better color image when compared to treating interpolation and fusion separately.     

Color enhancement of highly correlated images. I. Decorrelation and HSI contrast stretches

Conventional enhancements for the color display of multispectral images are based on independent contrast modifications or “stretches” of three input images. This approach is not effective if the image channels are highly correlated or if the image histograms are strongly bimodal or more complex. Any of several procedures that tend to “stretch” color saturation while leaving hue unchanged may better utilize the full range of colors for the display of image information. Two conceptually different enhancements are discussed: the “decorrelation stretch”, based on principal-component (PC) analysis, and the “stretch” of “hue”-“saturation”-intensity (HSI) transformed data. The PC transformation is scene-dependent, but the HSI transformation is invariant. Examples of images enhanced by conventional linear stretches, decorrelation stretch, and by stretches of HSI transformed data are compared. Schematic variation diagrams or two- and three-dimensional histograms are used to illustrate the “decorrelation stretch” method and the effect of the different enhancements.     

Parallel probabilistic relaxation labelling based on Markov random fields for spectral-spatial hyperspectral image classification

The large volume of data and computational complexity of algorithms limit the application of hyperspectral image classification to real-time operations. This work addresses the use of different parallel processing techniques to speed up the Markov random field (MRF)-based method to perform spectral-spatial classification of hyperspectral imagery. The Metropolis relaxation labelling approach is modified to take advantage of multi-core central processing units (CPUs) and to adapt it to massively parallel processing systems like graphics processing units (GPUs). The experiments on different hyperspectral data sets revealed that the implementation approach has a huge impact on the execution time of the algorithm. The results demonstrated that the modified MRF algorithm produced classification accuracy similar to conventional methods with greatly improved computational performance. With modern multi-core CPUs, good computational speed-up can be achieved even without additional hardware support. The CPU-GPU hybrid framework rendered the otherwise computationally expensive approach suitable for time-constrained applications.     

融合弱监督目标定位的细粒度小样本学习

目的 小样本学习旨在通过一幅或几幅图像来学习全新的类别。目前许多小样本学习方法基于图像的全局表征,可以很好地实现常规小样本图像分类任务。但是,细粒度图像分类需要依赖局部的图像特征,而基于全局表征的方法无法有效地获取图像的局部特征,导致很多小样本学习方法不能很好地处理细粒度小样本图像分类问题。为此,提出一种融合弱监督目标定位的细粒度小样本学习方法。方法 在数据量有限的情况下,目标定位是一个有效的方法,能直接提供最具区分性的区域。受此启发,提出了一个基于自注意力的互补定位模块来实现弱监督目标定位,生成筛选掩膜进行特征描述子的筛选。基于筛选的特征描述子,设计了一种语义对齐距离来度量图像最具区分性区域的相关性,进而完成细粒度小样本图像分类。结果 在mini Image Net数据集上,本文方法在1-shot和5-shot下的分类精度相较性能第2的方法高出0.56%和5.02%。在细粒度数据集Stanford Dogs和Stanford Cars数据集上,本文方法在1-shot和5-shot下的分类精度相较性能第2的方法分别提高了4.18%,7.49%和16.13,5.17%。在CUB 200-...     

基于HSI颜色空间统计直方图的图像检索

给出了一种对图像的特征进行提取、分析并识别出一定形状及色彩差别的方法,主要目的是提取图像的颜色变化。首先是对采集到的彩色图像进行预处理,包括:图像的滤波、目标图像的定位等;然后采取了相应的图像颜色空间和识别算法,包括:图像颜色空间的变换、颜色量化、图像特征提取、识别算法。计算机仿真结果表明,该算法可行,并取得了较好的效果。     

基于HSI颜色空间统计直方图的图像检索

给出了一种对图像的特征进行提取、分析并识别出一定形状及色彩差别的方法,主要目的是提取图像的颜色变化。首先是对采集到的彩色图像进行预处理．包括：图像的滤波、目标图像的定位等;然后采取了相应的图像颜色空间和识别算法,包括：图像颜色空间的变换、颜色量化、图像特征提取、识别算法。计算机仿真结果表明,该算法可行,并取得了较好的效果。     

Statistical skin color detection method without color transformation for real-time surveillance systems

Skin color is the significant information for many emerging applications in surveillance systems. However, the common skin color models usually need to perform color space transformation. This is not suitable for direct hardware implementation. This paper develops a statistical skin color model using the default RGB color space, which is especially suitable to implement on hardware for image processing applications. Moreover, an efficient face detection system is also proposed with our skin color model for hardware implementation. Compared with other skin color models, the proposed model produces the highest detection rate. Furthermore, the extended face detection system also significantly decreases the computational cost of the hardware implementation based on our skin color model. Experimental results demonstrate that our proposed detection system can be easily implemented on a field-programmable gate array (FPGA), where only 3202 logic cells is occupied with the high detection rate.