Generating B-Scans of the Environment With a Conventional Sonar

This paper examines the task of displaying more information about the environment using a conventional ranging sonar than is available in standard time-of-flight (TOF) maps. A conventional ranging sonar forms an environmental image that displays the information in the echo envelope, similar to a medical ultrasound image. The sonar performs rotational sector scans of simple objects and two complex environments containing various reflecting structures. In acquiring sonar data, we repeatedly reset the conventional sonar to generate a point process whose density relates to the echo amplitude. This point process is displayed as a grayscale image, called a brightness scan (B-scan), analogous to B-scans in medical ultrasound. We compare the information content of sonar B-scans to TOF maps for object classification and show B-scans to be richer. B-scan textures produced by rough surfaces and volumes containing random scatterers exhibit statistical invariance, similar to some organs within the body, suggesting the feasibility of automated classification. Image artifacts and means for their identification are discussed. The qualitative information present in sonar B-scans should lead to improved quantitative techniques for classifying objects.     

Modified gamma correction method to enhance ball grid array image for surface defect inspection

BGA (Ball Grid Array) surface defect detection requires faster and more accurate methods for semiconductor industry applications. Traditionally, the BGA inspection used gray-scale images. However, the solder pad, wiring and gray scales shown in images depict little variance. Therefore, when the threshold value is poorly set or the contract rate is insignificant, BGA detection may fail to segment an object. This research proposes a modified methodology that uses Gamma correction for image enhancement. Three-color bands were applied to a modified Gamma correction algorithm (i.e. RGB) to better separate the high and low image contrasts. Better results were obtained by dividing the image into background and foreground portions using the Gamma correction. As a result, the proposed method improved the contrast value by 52.09%. After the images were enhanced and segmented, the compactness and internal holes were calculated as features for classification. The results showed that classification correctness was 96.43%. The proposed method used a 640?×?480 pixel image, performing complete defect detection 0.3 seconds faster than the traditional enhancement method, which requires 1?second. The research results provide an effective solution for the detection and classification of the BGA surface tin ball defect problem.     

舰载声呐与航空吊放声呐协同双基地搜索范围研究

针对舰载声呐与航空吊放声呐构成的异类传感器双基地工作问题,建立了基于声呐方程的海洋环境噪声背景下双基地搜索范围模型。在舰载声呐与航空吊放声呐双基地协同工作系统中,舰载声呐发射探测信号,舰载声呐和航空吊放声呐均接收水下目标回波信号。根据能量关系建立了扩展面积、扩展距离、等效半径模型,定量分析了声源级、基线长度等因素对系统搜索范围的影响。仿真结果表明,当舰载声呐声源级相对于吊放声呐声源级足够大时,协同双基地工作模式在较大的基线长度范围内不仅在搜索范围上优于舰载声呐和吊放声呐均独立工作时的单基地联合搜潜,而且在搜索宽度上也有优势。     

基于修正CycleGAN的声呐图像库构建方法研究

声呐图像数据集获取困难,导致很多水下工作无法正常开展,如水下目标检测与跟踪、声呐图像的超分辨等,因此构建充足的声呐图像数据库成为很多水下研究工作的重要前提条件。受光学图像与合成孔径雷达（Synthetic Aperture Radar,SAR）图像转换研究工作的启发,提出了基于CycleGAN实现声呐图像库的构建,即利用光学图像合成声呐图像,实现光学到声呐的图像风格迁移。通过对CycleGAN网络损失函数的改进,提高了声呐图像的合成效果。通过与Pix2Pix等图像风格迁移网络进行比较的实验结果证明,修正后的CycleGAN网络具有更好的图像风格迁移效果。最后用合成的声呐图像训练Mask RCNN目标检测网络,并用真实的声呐图像进行测试,训练得到的模型能够成功地检测出真实声呐图像中对应的目标,进一步验证了利用光学图像构建声呐图像库的有效性。     

基于粒子群最小二乘支持向量机的前视声呐目标识别

随着声成像技术的日益发展和广泛应用,利用图像声呐进行水下目标识别逐渐成为水声探测领域的重要研究方向之一。根据前视声呐图像的特性,提出了一种水下目标识别的方法。对声呐图像进行去噪和增强处理并分割图像,来获取目标所在区域、提取目标的区域形状特征;利用粒子群算法优化最小二乘支持向量机的正则化参数和核参数,构造出高性能的多分类器;输入待识别目标的特征实现分类。实验表明:优化后的最小二乘支持向量机能够准确、有效地识别出水下目标,并且具有较高的精度。     

基于多波束测深声呐的海底气体泄漏实时报警系统

随着人们对海底油气资源的开发,在海底铺设的管道越来越多,海底管道泄漏风险日益增加,亟需高效、实时的检测方法。文章利用自主式水下航行器(Autonomous Underwater Vehicle, AUV)平台,搭载多波束测深声呐与基于VxWorks的嵌入式处理单元,设计了一套水下气体泄漏自动检测系统,将多波束测深声呐采集的水体图像数据传送给嵌入式处理单元,采用模板匹配方法对图像数据进行处理,完成对海底管道气体泄漏点的实时检测。湖试和海试都验证了该系统方案的有效性与先进性。     

A 3.2 kHz, 14-Bit Optical Absolute Rotary Encoder With a CMOS Profile Sensor

We have developed a 3.2 kHz, 14-bit optical absolute rotary encoder system using a profile sensor and a slit disc. The profile sensor, which we have designed for this purpose, is a unique CMOS area image sensor aimed at high-speed position detection of X and Y axes. Y axis profile data from the profile sensor is used for recognition of an index code on the slit disc, and X axis profile data is used for position detection of the index code. This combination of two-axis information enables high resolution even with a small number of index codes (64 indexes) and a small amount of digital calculations. Experimental results show that this system can detect rotation with an angular resolution of 14 bits (0.022deg) and a maximum detection speed of 3.2 kHz.     

Model-based sea mine classification with synthetic aperture sonar

The quality and effectiveness of sensor information provided by mine-hunting autonomous underwater vehicles (AUVs) equipped with high-resolution sonars has improved drastically in recent years. In parallel, data rates have significantly increased resulting in information overload. Automatic target recognition (ATR) is regarded as a solution for this problem. This study describes a specific ATR technique based on model matching for application to high-resolution data. A sonar model for generation of high-resolution synthetic aperture sonar (SAS) images is described and applied both as database generator and classification. The performance of the model matching, which is attained by correlation and stochastically, is evaluated using a large data set covering the variety expected in mine-hunting operations. The model-based features generated in this way are able to reach an acceptable classification performance. The article is concluded with one real data example, which is easily classified when training with the simulated database. Further work is next aimed to confirm performance on real data.     

基于背景抑制与改进多尺度LSD的声呐小目标检测算法

针对声呐小目标检测由于水下环境复杂、目标回波信号弱等因素造成虚警率和误检率较高的问题,文章提出基于背景抑制和改进直线分割检测(Line Segment Detection, LSD)的检测算法。首先对原始声呐数据截取序列片段,构建多周期累积历程图,凸显运动目标轨迹线特征;其次设计边缘滤波算子,有效滤除部分背景噪声,并结合投影变换进行线特征增强,不仅实现了断裂直线重连,还抑制了剩余噪声;然后基于图像金字塔改进了多尺度LSD直线分割检测算法,有效缓解了过检测问题,大幅增加了直线平均长度;最后为了合并冗余检测信息,利用运动轨迹时空一致性特征设计后处理模块,提高了检测定位精度。通过多组无人遥控潜水器(Remotely Operated Vehicle, ROV)、潜水员、空心球靶小目标序列的湖试、海试数据的定量与可视化结果定性分析,实验结果显示,文中算法与传统LSD相比,误检率和漏检率分别降低了11.2和3.9个百分点,定位误差下降了1.495个像素。结果表明,文中所提算法大幅提高了声呐小目标检测精度,为后续水下目标识别、跟踪等任务奠定重要基础。     

基于声呐历程累积图像的弱回波目标检测方法

针对浅海随机噪声与混响背景下蛙人等弱回波强度、慢速小目标的检测问题,提出一种基于声呐历程累积图像的目标检测方法。首先根据声呐图像时域、空域相关性,采用背景空时归一化处理技术,抑制声呐背景中的静态混响、突发性噪声等强回波干扰。声呐历程累积图像集成了多帧声呐图像的信息,目标回波亮点由于运动连续性形成亮线特征,利用该特征,采用Radon恒虚警率(Radon Constant False Alarm Rate,Radon-CFAR)检测声呐历程累积图像中的目标短时运动轨迹,能够检测到低信噪比的目标。分析了空时归一化处理和检测算法的性能,并通过海试数据验证了该算法的有效性,可以检测到低信噪比的蛙人目标回波。     

基于声呐历程累积图像的弱回波目标检测方法

针对浅海随机噪声与混响背景下蛙人等弱回波强度、慢速小目标的检测问题,提出一种基于声呐历程累积图像的目标检测方法.首先根据声呐图像时域、空域相关性,采用背景空时归一化处理技术,抑制声呐背景中的静态混响、突发性噪声等强回波干扰.声呐历程累积图像集成了多帧声呐图像的信息,目标回波亮点由于运动连续性形成亮线特征,利用该特征,采...     

热轧重轨表面缺陷在线检测识别的关键技术研究

目前热轧重轨表面缺陷检测速度慢、精度低。为此,提出了一种基于机器视觉的热轧重轨表面缺陷在线检测系统。分析了过暗过曝区域交叠融合法与图像像素线互相关校验法两种方法提取特征缺陷等关键技术,并对模糊脉冲神经网络的表面缺陷分类效果进行了研究。实际应用证明,采用上述机器视觉的检测关键技术对热轧重轨表面进行缺陷检测识别,较大提高了检测速度和精度,且检测正确率在90%以上。     

Model-based approach to the detection and classification of mines in sidescan sonar

This paper presents a model-based approach to mine detection and classification by use of sidescan sonar. Advances in autonomous underwater vehicle technology have increased the interest in automatic target recognition systems in an effort to automate a process that is currently carried out by a human operator. Current automated systems generally require training and thus produce poor results when the test data set is different from the training set. This has led to research into unsupervised systems, which are able to cope with the large variability in conditions and terrains seen in sidescan imagery. The system presented in this paper first detects possible minelike objects using a Markov random field model, which operates well on noisy images, such as sidescan, and allows a priori information to be included through the use of priors. The highlight and shadow regions of the object are then extracted with a cooperating statistical snake, which assumes these regions are statistically separate from the background. Finally, a classification decision is made using Dempster-Shafer theory, where the extracted features are compared with synthetic realizations generated with a sidescan sonar simulator model. Results for the entire process are shown on real sidescan sonar data. Similarities between the sidescan sonar and synthetic aperture radar (SAR) imaging processes ensure that the approach outlined here could be made applied to SAR image analysis.     

基于机器视觉的柔性包装袋喷码缺陷检测研究

目的 针对传统喷码检测方法计算量大、字符区域定位不显著、识别准确率较低等不足,提出一种基于机器视觉的柔性包装袋喷码缺陷检测方法。方法 以柔性包装袋上喷码图像为研究对象,以滤波抑噪、阈值处理等技术对图像进行预处理,运用YOLO-V3网络模型对字符区域进行定位,并采用阈值和非极大值抑制算法提高喷码区域定位的显著性,通过改进AlexNet网络结构、运用多特征融合运算等方法,获取更为丰富的图像卷积特征,实现字符串的整体识别,从而提高喷码缺陷识别的准确率。结果 将YOLO-V3联合改进AlexNet的检测方法与传统喷码检测方法进行对比,结果表明,所设计喷码缺陷检测方法的分类准确率达到99.39%。结论 基于机器视觉的柔性包装袋喷码缺陷检测方法在模型计算量、字符区域定位显著性和字符识别准确率都有一定的优势,并有效解决了字符串整体识别的问题。     

基于纹理特征的合成孔径声纳图像目标检测研究

随着成像声纳技术的发展,声纳图像的目标检测与识别逐渐成为数字图像处理领域的一个重要研究课题。合成孔径声纳图像含有丰富的纹理特征,而灰度共生矩阵具有丰富的特征参数,可以从不同的角度对纹理进行细致刻画。采用灰度共生矩阵可以描述合成孔径声纳图像纹理方面的特征,通过计算灰度共生矩阵在方位向和距离向的能量、相关性、对比度和熵值,并构造特征向量,可以对合成孔径声纳图像中的目标进行准确检测。从实验结果可以看出,基于纹理信息可以准确实现合成孔径声纳图像的目标检测。     

Resampling Factor Estimation via Dual-Stream Convolutional Neural Network

The estimation of image resampling factors is an important problem in image forensics. Among all the resampling factor estimation methods, spectrumbased methods are one of the most widely used methods and have attracted a lot of research interest. However, because of inherent ambiguity, spectrum-based methods fail to discriminate upscale and downscale operations without any prior information. In general, the application of resampling leaves detectable traces in both spatial domain and frequency domain of a resampled image. Firstly, the resampling process will introduce correlations between neighboring pixels. In this case, a set of periodic pixels that are correlated to their neighbors can be found in a resampled image. Secondly, the resampled image has distinct and strong peaks on spectrum while the spectrum of original image has no clear peaks. Hence, in this paper, we propose a dual-stream convolutional neural network for image resampling factors estimation. One of the two streams is gray stream whose purpose is to extract resampling traces features directly from the rescaled images. The other is frequency stream that discovers the differences of spectrum between rescaled and original images. The features from two streams are then fused to construct a feature representation including the resampling traces left in spatial and frequency domain, which is later fed into softmax layer for resampling factor estimation. Experimental results show that the proposed method is effective on resampling factor estimation and outperforms some CNN-based methods.     

Monte Carlo simulation of the atmospheric point-spread function with an application to correction for the adjacency effect

Monte Carlo techniques are used to simulate atmospheric point-spread functions (PSF's) that are appropriate for the viewing geometries typical of the Airborne Visible-Infrared Imaging Spectrometer (AVIRIS). A model sensor is located at an altitude of 20 km and views a Lambertian surface through a horizontally homogeneous and vertically stratified atmosphere. Simulations show the effects on the PSF of variation of the aerosol phase function, the aerosol optical thickness, the sensor viewing angle, and the wavelength. An algorithm that uses the PSF to correct high-contrast images for adjacency effects is developed and applied to an AVIRIS image of Big Pine Key in the Florida Keys. A method to approximate the atmospheric PSF's without the need to resort to a Monte Carlo simulation is described. Correction of the AVIRIS image through the use of the approximated PSF is consistent with a previous correction. Error analysis is difficult and scene dependent; however, the correction algorithm is shown to be capable of indicating regions of high-contrast images in which conventional estimates of surface-leaving radiance are likely to be unreliable due to adjacency effects.     

基于声呐图像序列SIFT特征的多目标跟踪方法研究

针对高频成像声呐多目标跟踪中声成像的不稳定问题,在基于检测前跟踪的声呐图像序列尺度不变特征变换（Scale Invariant Feature Transform,SIFT）的基础上,提出了一种结合对数变换的声呐图像序列SIFT特征跟踪方法。该方法利用对数变换进行图像预处理,进而利用声呐图像序列SIFT特征进行多目标跟踪。水池试验数据的处理结果表明：与中值滤波和动态亮度分配的方法相比,结合对数变换的声呐图像序列SIFT特征跟踪方法具有更好的多目标跟踪效果。