Semantic Segmentation by Using Down-Sampling and Subpixel Convolution: DSSC-UNet

Recently, semantic segmentation has been widely applied toimage processing, scene understanding, and many others. Especially, indeep learning-based semantic segmentation, the U-Net with convolutionalencoder-decoder architecture is a representative model which is proposed forimage segmentation in the biomedical field. It used max pooling operationfor reducing the size of image and making noise robust. However, instead ofreducing the complexity of the model, max pooling has the disadvantageof omitting some information about the image in reducing it. So, thispaper used two diagonal elements of down-sampling operation instead ofit. We think that the down-sampling feature maps have more informationintrinsically than max pooling feature maps because of keeping the Nyquisttheorem and extracting the latent information from them. In addition,this paper used two other diagonal elements for the skip connection. Indecoding, we used Subpixel Convolution rather than transposed convolutionto efficiently decode the encoded feature maps. Including all the ideas, thispaper proposed the new encoder-decoder model called Down-Sampling andSubpixel Convolution U-Net (DSSC-UNet). To prove the better performanceof the proposed model, this paper measured the performance of the UNetand DSSC-UNet on the Cityscapes. As a result, DSSC-UNet achieved89.6% Mean Intersection OverUnion (Mean-IoU) andU-Net achieved 85.6%Mean-IoU, confirming that DSSC-UNet achieved better performance.

Analyzing the Effect of the Intra-Pixel Position of Small PSFs for Optimizing the PL of Optical Subpixel Localization

Subpixel localization techniques for estimating the positions of point-like images captured by pixelated image sensors have been widely used in diverse optical measurement fields.With unavoidable imaging noise,there is a precision limit(PL)when estimating the target positions on image sensors,which depends on the detected photon count,noise,point spread function(PSF)radius,and PSF’s intra-pixel position.Previous studies have clearly reported the effects of the first three parameters on the PL but have neglected the intra-pixel position information.Here,we develop a localization PL analysis framework for revealing the effect of the intra-pixel position of small PSFs.To accurately estimate the PL in practical applications,we provide effective PSF(e PSF)modeling approaches and apply the Cramér–Rao lower bound.Based on the characteristics of small PSFs,we first derive simplified equations for finding the best PL and the best intra-pixel region for an arbitrary small PSF;we then verify these equations on real PSFs.Next,we use the typical Gaussian PSF to perform a further analysis and find that the final optimum of the PL is achieved at the pixel boundaries when the Gaussian radius is as small as possible,indicating that the optimum is ultimately limited by light diffraction.Finally,we apply the maximum likelihood method.Its combination with e PSF modeling allows us to successfully reach the PL in experiments,making the above theoretical analysis effective.This work provides a new perspective on combining image sensor position control with PSF engineering to make full use of information theory,thereby paving the way for thoroughly understanding and achieving the final optimum of the PL in optical localization.

Ulva prolifera subpixel mapping with multiple-feature decision fusion

The unavoidable nature of Ulva prolifera mixed pixel in low-resolution remote sensing images would result in rough boundary of U.prolifera patches,omission of tiny patches,and overestimation of coverage area.The decomposition of U.prolifera mixed pixel addresses the issue of coverage area overestimation,and the remaining problems can be alleviated by subpixel mapping(SPM).Due to the drift and dissipation of U.prolifera,a suitable SPM method is the single image-based unsupervised method.However,the method has difficulties in detail reconstruction,insufficient learning of spectral information,and SPM error introduced by abundance deviation.Therefore,we proposed a multiple-feature decision fusion SPM(MFDFSPM)method.It involves three branches to obtain the spatial,abundance,and spectral features of U.prolifera while considers multi-feature information using the fusion strategy.Experiments on the Geostationary Ocean Color Imager images in the Yellow Sea of China indicate that the MFDFSPM overperforms several typical U.prolifera SPM methods in higher accuracy and stronger robustness in both SPM and abundance calculation,which produced subpixel map with more detailed spatial information and less noise.

Subpixel accuracy for extracting groove center based on corner detection

Subpixel accuracy for V-groove center in robot welding is researched and a software measure to increase the accuracy of seam tracking by laser is presented. LOG （ Laplacian of Gaussian ） operator is adopted to detect image edge. Vgroove center is extracted by corner detection of extremum curvature. Subpixel position is obtained by Lagarange polynomial interpolation algorithm. Experiment results show that the method is brief and applied, and is sufficient for the real time of robot welding by laser sensors.

Coastal Biomass Mapping Through Subpixel Classification Technique—A Technique of Segregating Materials Within Pixel

World Wide Fund for Nature(WWF)scientists in collaboration with regional experts around the world, have developed the Global 200 Ecoregions Program, the first comparative analysis of biodiversity to cover biologically outstanding terrestrial,freshwater and marine habitats,spanning five continents and all the world’s oceans.Due to its immense importance,Indus Delta Ecoregion is one of the G200 Ecoregions of

Subpixel Reconstruction Algorithm in Staring FPA Nonuniform Microscan Imaging

For the nonuniform microscan system where the interframe translation is no longer equivalent to accurate halfpixel, a 2-dimension non-interpolated subpixel algorithm is proposed to consider arhitrary value of microscanning. The aim of the proposed algorithm is to restore double resolution from 2 × 2 undersampled frames. To solve the ill-posed problem in the process of image reconstruction, the prior boundary condition is introduced, and the proposed subpixel reconstruction algorithm is reformulated into the form of line-by-line backward propagation iteration for reducing the calculation complexity. Since the direction of movement offset relative to the accurate halfpixel determines the transfer matrix of the image degradation process, the recon- struction is classified into 4 types when 2 × 2 mieroscan model is applied. All the simulation results and experiment data demonstrate the double resolution improvement compared with the undersampled images. The focus problem, scarcely any possibility of the operation with accurate halfpixel micromotion, is figured out for en-hancing the feasibility of subpixel reconstruction used in practice.

A New Edge-directed Subpixel Edge Localization Method

Localization of the inspected chip image is one of the key problems with machine vision aided surface mount devices (SMD) and other micro-electronic equipments. This paper presents a new edge-directed subpixel edge localization method. The image is divided into two regions, edge and non-edge, using edge detection to emphasize the edge feature. Since the edges of the chip image are straight, they have straight-line characteristics locally and globally. First, the line segments of the straight edge are located to subpixel precision, according to their local straight properties, in a 3×3 neighborhood of the edge region. Second, the subpixel midpoints of the line segments are computed. Finally, the straight edge is fitted using the midpoints and the least square method, according to its global straight property in the entire edge region. In this way, the edge is located to subpixel precision. While fitting the edge, the irregular points are eliminated by the angles of the line segments to improve the precision. We can also distinguish different edges and their intersections using the angles of the line segments and distances between the edge points, then give the vectorial result of the image edge with high precision.

Subpixel Target Location Techniques for 3-D Coordinate Measuring System

the close photogrammetric 3-D coordinate measurement is a newmeasuring technology in the fields of the coordinate measurementmachine (CMM) in recent years. In this method, we usually place sometargets on the measured object and take image of targets to determinethe object coordinate. The subpixel location of target image plays animportant role in high accuracy 3-D coordinate measuring procedure.In this paper, some subpixel location methods are reviewed and somefactors which affect location precision are analyzed.

A Novel Subpixel Curved Edge Localization Method

With the high-speed development of digital image processing technology, machine vision technology has been widely used in automatic detection of industrial products. A large amount of products can be treated by computer instead of human in a shorter time. In the process of automatic detection, edge detection is one of the most commonly used methods. But with the increasing demand for detection precision,traditional pixel-level methods are difficult to meet the requirement, and more subpixel level methods are in the use. This paper presents a new method to detect curved edge with high precision. First, the target area ratio of pixels near the edge is computed by using one-dimensional edge detection method. Second, parabola is used to approximately represent the curved edge. And we select appropriate parameters to obtain accurate results. This method is able to detect curved edges in subpixel level, and shows its practical effectiveness in automatic measure of products with arc shape in large industrial scene.

基于多时相光学遥感影像亚像素相关的边坡监测——以溪洛渡电站为例

水电站高边坡监测是水电站灾害防治中的关键问题,基于卫星遥感技术进行库区滑坡监测是解决该问题的重要手段之一。利用卫星影像亚像素相关性算法获取地表形变位移场,能够克服SAR影像失相关因素,在水电站边坡监测中具有重大的应用潜力。本文以溪洛渡电站2015-2019年5期Google Earth影像和2019-2022年4期Sentinel-2影像为数据源,采用相位相关算法计算了边坡形变量,通过构建一次多项式曲面拟合模型去除轨道误差等趋势性误差,获取了2016、2017和2019年溪洛渡电站下游边坡形变值。分析显示,两种影像提取得到的边坡形变量具有相同的变化趋势,均与谷幅实测数据吻合较好。本文结果验证了基于多时相遥感数据将亚像素相关性匹配技术运用于大型水电站边坡形变监测的可行性。

基于实际工程项目的装配参数视觉检测实践教学设计

根据“新工科”建设对实践教学的要求,结合精密制造行业视觉检测实际工程项目,对智能制造专业的机器视觉课程实践教学进行了教学案例设计。根据实际工程项目案例将教学实验划分为基础型、综合型、创新型三类,旨在增强学生的工程体验感,使学生能够循序渐进地掌握机器视觉的知识点和应用技能,促进本科实践教学课程建设,提高学生的创新能力和工程实践能力。

焊缝缺陷三维成像及亚像素定量方法研究

为实现工业设备中焊缝结构内部缺陷准确地定量分析,针对构造缺陷三维重构模型时缺陷图谱的过分割和欠分割导致定量分析不够精确的问题,提出了一种基于图谱颜色分割的新算法。利用该算法处理缺陷切片并进行三维重构获取其长度,应用Canny算子与三次样条插值法结合获取缺陷亚像素边缘,基于缺陷区域连通标记算法和累加法可获取缺陷面积和体积的方法。研究结果表明:与实际数据对比,未焊透、未熔合、夹渣缺陷重构的长度误差分别为8.18%、5%、9.62%,基于此对分割缺陷的亚像素边缘内连通域进行标记以准确地获取其面积及体积。该方法能够在不破坏焊缝结构的情况下即时获取焊缝内部缺陷的形貌并对其进行精准且全面的定量分析。

基于自适应ROI的轴承内圈高精度视觉尺寸检测

针对机器视觉轴承内圈侧面复杂形状尺寸检测精度低的问题,提出根据检测目标建立小面积感兴趣区域(Region of Interest,ROI)的自适应选取方法和基于Zernike矩的ROI亚像素级边缘提取方法,大幅提升了轴承内圈尺寸的检测精度。首先分别拍摄轴承内圈左侧与右侧轮廓图像,对图像进行预处理。在此基础上,通过角点检测融合像素扫描的方法实现自适应ROI选取,解决了因轴承内圈移动引起的小面积ROI边缘误判问题;使用Canny算子提取ROI的像素级边缘,再用改进的Zernike矩算法得到亚像素级边缘。最后,分别对ROI中提取的边缘进行最小二乘圆拟合和直线拟合,根据像素当量与视场间隔将图像中各尺寸转换为轴承内圈实际尺寸。实验结果表明:所提方法测量的标准不确定度低于0.005 mm,满足轴承尺寸高精度检测的要求,对于实现轴承检测的自动化有实际意义。

架空导线压接对边距的图像测量方法

压接导线主要应用于架空线路建设,其对边距对线路安全运行及服役期限起到重要作用。针对传统人工测量存在的可靠性差、效率低的缺点,提出了一种基于图像处理的压接对边距测量方法。首先根据图像的累计灰度突变点自动裁剪被测区域并进行中值滤波,然后基于改进Otsu阈值法获得边缘连续的二值图像;再分别使用形态学处理和高斯曲线拟合法对边缘进行二次定位,得到亚像素级边缘点;最后采用基于随机抽样一致性(random sample consensus, RANSAC)的最小二乘拟合方法完成上下边缘点的曲线拟合,实现压接对边距的测量。通过模拟试验表明了RANSAC拟合经过10次迭代后测量算法的效果较好,测量误差小于0.1 pixels;而实际导线测量试验结果表明,改进Otsu算法对光照强度具有适应性,同时与人工测量相比,图像测量方法的最大相对偏差为1.82%,重复标准差提高60%,平均耗时仅为人工测量的1/10,能够实现压接对边距高效可靠的测量。

基于双目立体视觉的连铸辊尺寸三维测量方法

针对连铸车间中,扇形段连铸辊尺寸人工测量效率低下的问题,提出了一种基于双目视觉系统的连铸辊尺寸测量方法。首先,对双目相机采集的图像进行预处理并采用Otsu法分割工件前后背景;接着,针对边缘检测精度不高的问题,将传统Canny边缘检测算法的梯度模板增加到8个以提取工件轮廓,结合多项式插值公式提取亚像素级别特征点;然后,在SAD与Census变换融合的立体匹配基础上引入RANSAC算法来消除错误匹配;最后,采用三角测量原理计算出零件的尺寸。实验结果表明,系统测量的平均相对误差为0.14%,测量方法具有较高的精度,其稳定性与精确性满足连铸辊的尺寸自动检测任务。

融合结构光的折反射全景单目视觉测距

无人机在大机动运动时视场变化较大,在弱纹理环境下视觉特征易丢失,这降低了视觉距离测量的精度。针对此问题,本文将折反射全景相机与结构光相结合,进一步提出了融合结构光的折反射全景单目视觉测距方法。在本研究中,为了提高测距方法的精度以及处理速度,首先对结构光进行感兴趣区域(ROI)提取,并运用中值滤波对激光条纹图像进行去噪后,再采用改进的Steger算法进行亚像素级别的结构光光条中心坐标提取。其次,运用最小二乘原理对提取的亚像素点进行曲线拟合,得到结构光的曲线方程。最后,通过理论推导求取空间平面参数,建立测距目标函数,进一步求得距离信息。实验结果表明:在不同场景下,所提出测距方法的测距精度均在3%以内,证明了该方法的有效性。满足无人机在大机动运动条件下以及弱纹理场景中稳定可靠的测距需求,进一步提升了无人机环境感知能力。

基于双三次插值和高斯拟合的光条中心亚像素提取方法

针对结构光三维测量的高精度要求,提出一种基于双三次插值和高斯拟合的激光线条亚像素提取方法。该方法首先对光条图像进行感兴趣区域提取、高斯滤波等前处理,其次用形态学方法细化激光线条得到初始中心点,然后使用五点拟合法得到光条法线方向,在光条法线方向上取亚像素的拟合点,利用双三次插值得到亚像素坐标的像素值,最后进行高斯拟合,到光条亚像素中心。实验结果表明,算法平均精度值为0.389147像素,消耗时间为0.056 2 s,具有较高的提取精度和稳定性。

基于残差网络的字符编码标志点识别

针对大尺寸结构近景摄影测量中如何提高编码标志的容量和识别率等问题,设计了一种由中心圆,字母和数字组成的字符编码标志,提出了一种基于ResNet18模型改进的字符编码点识别方法和基于Zernike矩提出的亚像素定位方法。首先,利用均值滤波、自适应阈值处理和形态学中的闭运算等方法对图像进行预处理,根据连通域分析与筛选对预处理好的图像进行分割得到字符编码点;其次,用字符编码点识别网络进行识别,最终通过亚像素定位方法对编码点进行定位。从实验结果得出字符编码点的识别准确率可达99.89%。该类标志点定位精度高,提出的识别方法准确率高。

基于改进Canny和灰度矩的水下图像边缘检测方法研究

针对水下图像对比度差、模糊,致使检测出的边缘存在不连续和伪边缘的问题,提出一种基于改进的Canny和亚像素的水下图像边缘检测方法。采用改进的Canny算法检测水下图像的像素级边缘,在此基础上采用灰度矩提取图像的亚像素边缘特征,提高边缘的定位精度和检测率。实验结果表明,提出的算法相较于传统的像素级边缘检测算法在边缘轮廓的提取、减少伪边缘、提升边缘精度方面有较大优势,尤其是对深海环境中光源单一以及图像对比度差的物体边缘的检测具有更显著的效果和定位精度。

基于主动位移成像的图像超分辨率重建

超分辨重建算法是一种将低分辨率图像恢复为高分辨率图像的算法,被广泛用于医学、遥感、军事安防以及人脸识别等领域。在黑夜、远场场景下构建数据集比较困难,基于深度学习的超分辨重建算法应用受到阻碍。而微扫描成像技术扫描模式固定,对器件到位精度要求高。针对这两个问题,我们提出一种基于主动位移成像的图像超分辨率重建算法。具体地,在控制相机随机移动的同时记录采样时刻位移,通过解算、映射选图、精确匹配图像序列并获取多帧图像间的亚像素信息,然后对估计图像进行迭代和更新,最后重建获得高分辨率图像。实验结果表明,本算法在PSNR、SSIM和平均梯度三个指标上都优于最近提出的基于POCS的图像超分辨率重建算法MFPOCS,与基于CNN的方法ACNet相比具有竞争力。值得提出的是,本算法无需固定的扫描模式,降低了微扫描技术对器件实时到位精度的要求,同时,本算法可以保证重建初始帧的优良选取,有效规避了POCS算法的固有缺点。