结合双注意力机制的道路裂缝检测

目的 道路裂缝检测旨在识别和定位裂缝对象,是保障道路安全的关键问题之一。为解决传统深度神经网络在检测背景较复杂、干扰较大的裂缝图像时精度较低的问题,设计了一种基于双注意力机制的深度学习道路裂缝检测网络。方法 本文提出了在骨干网络中融入空洞卷积和两种注意力机制的方法,将其中的轻量型注意力机制与残差模块结合为残差注意力模块Res-A。对比研究了该模块“串联”和“并联”两种方式对于裂缝特征关系权重的影响并获得最佳连接。同时,引入Non-Local计算模式的注意力机制,通过挖掘特征图谱的关系权重以提高裂缝检测性能。结合两种注意力机制可以有效解决复杂背景下道路裂缝难检测的问题,提高了道路裂缝检测精度。结果 在公开复杂道路裂缝数据集Crack500上进行对比实验与验证。为证明本文网络的有效性,将平均交并比(mean intersection over union, m Io U)、像素精确度(pixel accuracy, PA)和训练迭代时间作为评价指标,并进行了3组对比实验。第1组实验用于评价残差注意力模块中通道注意力机制和空间注意力机制之间不同组合方式的检测性能,结果表明这两种机制并联相加时...     

Edge-preserving color image denoising through tensor voting

This paper presents a new method for edge-preserving color image denoising based on the tensor voting framework, a robust perceptual grouping technique used to extract salient information from noisy data. The tensor voting framework is adapted to encode color information through tensors in order to propagate them in a neighborhood by using a specific voting process. This voting process is specifically designed for edge-preserving color image denoising by taking into account perceptual color differences, region uniformity and edginess according to a set of intuitive perceptual criteria. Perceptual color differences are estimated by means of an optimized version of the CIEDE2000 formula, while uniformity and edginess are estimated by means of saliency maps obtained from the tensor voting process. Measurements of removed noise, edge preservation and undesirable introduced artifacts, additionally to visual inspection, show that the proposed method has a better performance than the state-of-the-art image denoising algorithms for images contaminated with CCD camera noise.     

Texture crack detection

Automatic visual inspection has become one of the active research issues in machine vision technology in the past decades. Most of the methodologies developed address the problems of defect detection on a nontextured or regularly textured surface. However, problems in detecting defects on a randomly textured surface, especially cracks, have not received much attention. In this paper, we present a novel algorithm that uses a Wigner model to identify cracks in complex textural backgrounds, regardless of whether the inspected surface is randomly or regularly textured. We also investigate the windowing characteristics of the Wigner distribution and their impact on crack detection. Some of the Brodatz' natural texture images have been used for evaluating the performance of the algorithm. Promising results are obtained and presented in this paper.     

基于可见光视觉图像的路面裂缝识别深度学习方法述评

基于可见光视觉图像的表面裂缝识别为非接触式,不受被测对象材质限制,可在线自动检测,具有速度快、成本低和精度高等优势。首先较为全面地搜集了典型的路面裂缝公开数据集,整理归纳了样本特征及其随机可变影响因素,并比较了传统手工设计特征工程、机器学习和深度学习3种主要裂缝识别方法的优缺点。然后,从网络架构、性能和效果方面着重评述了自搭架构、迁移学习和编码-解码器等易于训练和部署的深度学习算法新进展,通过算法优化和算力提升可显著提高识别的效果和性能,测试结果表明能够在低算力平台上实现裂缝补丁级快速检测和像素级实时检测。     

FoSA: F* Seed-growing Approach for crack-line detection from pavement images

Most existing approaches for pavement crack line detection implicitly assume that pavement cracks in images are with high contrast and good continuity. This assumption does not hold in pavement distress detection practice, where pavement cracks are often blurry and discontinuous due to particle materials of road surface, crack degradation, and unreliable crack shadows. To this end, we propose in this paper FoSA — F* Seed-growing Approach for automatic crack-line detection, which extends the F* algorithm in two aspects. It exploits a seed-growing strategy to remove the requirement that the start and end points should be set in advance. Moreover, it narrows the global searching space to the interested local space to improve its efficiency. Empirical study demonstrates the correctness, completeness and efficiency of FoSA.     

Pothole detection in asphalt pavement images

Pavement condition assessment is essential when developing road network maintenance programs. In practice, the data collection process is to a large extent automated. However, pavement distress detection (cracks, potholes, etc.) is mostly performed manually, which is labor-intensive and time-consuming. Existing methods either rely on complete 3D surface reconstruction, which comes along with high equipment and computation costs, or make use of acceleration data, which can only provide preliminary and rough condition surveys. In this paper we present a method for automated pothole detection in asphalt pavement images. In the proposed method an image is first segmented into defect and non-defect regions using histogram shape-based thresholding. Based on the geometric properties of a defect region the potential pothole shape is approximated utilizing morphological thinning and elliptic regression. Subsequently, the texture inside a potential defect shape is extracted and compared with the texture of the surrounding non-defect pavement in order to determine if the region of interest represents an actual pothole. This methodology has been implemented in a MATLAB prototype, trained and tested on 120 pavement images. The results show that this method can detect potholes in asphalt pavement images with reasonable accuracy.     

Feature detection of triangular meshes based on tensor voting theory

This paper presents n-dimensional feature recognition of triangular meshes that can handle both geometric properties and additional attributes such as color information of a physical object. Our method is based on a tensor voting technique for classifying features and integrates a clustering and region growing methodology for segmenting a mesh into sub-patches. We classify a feature into a corner, a sharp edge and a face. Then, finally we detect features via region merging and cleaning processes. Our feature detection shows good performance with efficiency for various dimensional models.     

SEDMI: Saliency based edge detection in multispectral images

Detecting edges in multispectral images is difficult because different spectral bands may contain different edges. Existing approaches calculate the edge strength of a pixel locally, based on the variation in intensity between this pixel and its neighbors. Thus, they often fail to detect the edges of objects embedded in background clutter or objects which appear in only some of the bands.We propose SEDMI, a method that aims to overcome this problem by considering the salient properties of edges in an image. Based on the observation that edges are rare events in the image, we recast the problem of edge detection into the problem of detecting events that have a small probability in a newly defined feature space. The feature space is constructed by the spatial gradient magnitude in all spectral channels. As edges are often confined to small, isolated clusters in this feature space, the edge strength of a pixel, or the confidence value that this pixel is an event with a small probability, can be calculated based on the size of the cluster to which it belongs.Experimental results on a number of multispectral data sets and a comparison with other methods demonstrate the robustness of the proposed method in detecting objects embedded in background clutter or appearing only in a few bands.     

沥青路面裂缝图像检测算法研究

针对沥青路面裂缝图像边缘提取不精确的问题,基于形态学多尺度的思想,结合沥青路面裂缝图像中裂缝形状不规则,裂缝信息比较弱小的特点,提出了一种沥青路面裂缝图像检测算法。算法构造了多结构元素的抗噪型边缘检测算子,且依据不同形状的结构元素对裂缝边缘填充的几率不同,设定了一种自适应确定权重的算法,使得此算子在平滑滤波的同时又检测到各种类型的边缘。与传统裂缝图像检测算法相比,该算法有效抑制了各种噪声的影响,实现了多种类型边缘的精确提取;理论分析和实验结果表明该算法是可行且有效的。     

CrackEmbed: Point feature embedding for crack segmentation from disaster site point clouds with anomaly detection

Laser-scanned point clouds can be used to represent the 3D as-damaged condition of building structures in a post-disaster scenario. Performing crack detection from the acquired point clouds is a critical component of disaster relief tasks such as structural damage assessment and risk assessment. Crack detection methods based on intensity or normals commonly result in noisy detections. On the other hand, deep learning methods can achieve higher accuracy but require a large dataset of annotated cracks. This research proposes an unsupervised learning framework based on anomaly detection to segment out cracked regions from disaster site point clouds. First, building components of interest are extracted from the point cloud scene using region growing segmentation. Next, a point-based deep neural network is used to extract discriminative point features using the geometry of the local point neighborhood. The neural network embedding, CrackEmbed, is trained using the triplet loss function on the S3DIS dataset. Then, an anomaly detection algorithm is used to separate out the points belonging to cracked regions based on the distribution of these point features. The proposed method was evaluated on laser-scanned point clouds from the 2015 Nepal earthquake as well as a disaster response training facility in the U.S. Evaluation results based on the point-level precision and recall metrics showed that CrackEmbed in conjunction with the isolation forest algorithm resulted in the best performance overall.     

基于张量投票的昆虫翅脉提取算法

昆虫翅脉提取对于昆虫自动分类意义重大。鉴于传统翅脉提取算法中存在断裂、边缘不整齐等缺点,提出一种基于张量投票的昆虫翅脉提取算法。首先对翅脉图像进行去噪、二值化、形态学等预处理,得到稀疏二值点图,然后计算每点的张量值,结合Gestalt定律的邻近性、相似性规则,对邻域内点进行张量投票,并设定投票阈值,最终获取昆虫翅脉轮廓。实验结果表明,由于引入张量和Gestalt规则,文中所提算法可以提取出更加符合感知规则的翅脉结构,得到较为完整且平滑的翅脉轮廓,同时对于出现少许断裂的翅脉图像,依然可以获得较为完整的翅脉边缘,这为后期的昆虫自动分类奠定基础。     

基于局部纹理特征的沥青路面裂缝检测方法

针对不均匀光照和阴影等因素影响沥青路面图像中裂缝检测误识别问题,提出一种基于局部纹理特征的沥青路面裂缝检测方法.设计结构保持型Retinex算法将高频的纹理信号从低频光照信号和结构型纹理中分离,改进百分比阈值算法,获取高信噪比的裂缝区域二值图像,建立高置信裂缝段的特征匹配机制,利用圆形度、面积、不同置信裂缝的类间欧氏距...     

Automatic detection of blood vessels from CTA images employing morphological operation

Recently, the disease arteriosclerosis has rapidly increased, and in particular the type that originates in the peripheral arteries of the hands and feet and is called arteriosclerosis obliterans (ASO). ASO is one of the typical diseases that cause chronic ischemia of the limbs and lead to obstruction of the blood flow. In diagnosing ASO, lower contrast enhanced computed tomography (CT) is useful for observing the arteries. However, it is a huge task for radiologists to separate accurately only the arteries which are involved in the disease, using the manual or semiautomatic proprietary software which is supplied for lower computed tomographic angiography (CTA). We have developed a new technique for the detection of arteries from CTA images by the use of a morphological operation. In this technique an N-Quoit filter, which is a useful filter for the detection of lung nodules, is applied to identify the arteries in CTA images. Some experimental results have shown good performance in the segmentation of arteries. This work was presented in part at the 13th International Symposium on Artificial Life and Robotics, Oita, Japan, January 31–February 2, 2008     

视网膜血管图像特征点自动提取和分类

由动静脉血管组成的眼底视网膜血管结构的特征点是预测心血管疾病、图像分析和生物学应用的重要特征。把角点检测引入到视网膜血管分叉点和交叉点提取中,利用边缘检测算子得到二值边缘图像,采用基于累加点到弦的距离（CPDA）的角点检测方法得到候选特征点,再根据视网膜血管图像的拓扑结构设计自适应矩形探测器对候选特征点进行删减和分类。实验结果表明,基于CPDA的角点检测和自适应矩形探测器的方法有效地实现了节点的提取和分类。     

多尺度局部特征增强Transformer道路裂缝检测模型

目的 道路裂缝是路面病害的早期征兆。定期监测路面状况、及时准确地发现路面裂缝对于交通养护机构降低成本、保证路面结构的可靠性和耐久性以及提高驾驶安全性、舒适性有重要意义。目前基于卷积神经网络的深度学习模型在长距离依赖建模方面存在不足,模型精度难以满足真实路面环境下的裂缝检测任务。一些模型通过引入空间/通道注意力机制进行长距离依赖建模,但是会导致计算量和计算复杂程度增加,无法实现实时检测。鉴于此,本文提出一种基于Transformer编码—解码结构的深度神经网络道路裂缝检测模型CTNet(crack transformer network)。方法 该模型主要由Transformer注意力模块、多尺度局部特征增强模块、上采样模块和跨越连接4部分构成。采用Transformer注意力机制能更有效提取全局和长距离依赖关系,克服传统卷积神经网络表征输入信息的短距离相关缺陷。同时,为适应裂缝尺寸变化多样性,将Transformer与多尺度局部特征增强模块相结合,从而有效整合不同尺度局部信息,克服Transformer局部特征建模不足。结果 通过与DeepCrack模型在不同裂缝检测数据集中的比较表明...     

Gradient-based edge detection and feature classification of sea-surface images of the Southern California Bight

An edge detection algorithm was developed that is capable of objectively detecting significant edges in remotely sensed images of the surface ocean. The algorithm utilizes a gradient-based edge detector that is less sensitive to noise in the input image than previously used detectors and has the ability to detect edges at different length scales. The algorithm was used to provide a statistical view of surface front occurrences in the Southern California Bight using six years of satellite observations of sea-surface temperature (AVHRR) and chlorophyll (SeaWiFS). Regions of high front occurrence probability were identified near capes, headlands, and islands. In onshore direction, chlorophyll concentration increased and temperature decreased, indicative of coastal upwelling. The algorithm was further applied to coincident time series of temperature and chlorophyll to investigate the event-scale dynamics of mesoscale features and the spatial relationships between physical and biological processes. A simple scheme for identifying and classifying eddies delineated by the edge detection algorithm was developed to yield a census of eddy occurrence in the Southern California Bight.     

Generating fuzzy edge images from gradient magnitudes

The representation and processing of edges in images based on notions from fuzzy set theory has become popular in recent years. There are several reasons for this direction, from the vague definition of edges to the inherent uncertainty of digital images. Here, we study the transition from a gradient image, a popular intermediate representation, to a fuzzy edge image. We consider different parametric membership functions to transform the gradients into membership degrees. A histogram-based strategy is then introduced for automatically determining the value of those parameters, adapting the membership functions to the characteristics of each image. The functions are applied on the Canny method for edge detection, resulting in an improvement compared to the classical normalizing approach.     

面向线阵列CCD道路影像的裂缝识别

目的 道路裂缝的等级评定是公路养护的基本任务之一,目前国内相关部门主要通过线阵列相机采集道路影像,由于道路影像裂缝的识别会受到多种因素干扰（树木及车辆的投影、光照变化、油渍、树枝与稻草等条状物、各类垃圾）,降低了基于道路影像自动识别裂缝算法的准确率,导致对于路面等级评价依旧采用人工的方式进行,为此提出一种道路影像裂缝鲁棒识别方法。方法 由于采集的图像尺寸较大,同时为了避免光照不均匀带来的问题,首先对图像进行分块,采用改进的CV模型对分块影像进行预处理,获得初步的分割结果。其次,通过以下4个特点识别线阵列CCD道路影像的裂缝：1）裂缝在分块区域中占据较小的面积比;2）裂缝在影像中呈现的连续性较差;3）裂缝的宽度与长度比值较小;4）同一段裂缝的走向基本一致。为了利用裂缝的后两项特点,采用椭圆拟合的方法计算初步分割区域的方向,并以此为基础将这些区域分为4类。在每个分类中,分别计算各区域内的质心位置,建立质心间的矢量表,设计递归算法计算其共线性,获得裂缝检测结果,并以此为基础构造活动模型的初始距离矩阵,通过在原图中迭代求解更为精确的裂缝区域。结果 从2 000幅道路影像中挑选包含道路裂缝的影像100幅,并按序号等间隔分别取出5组未含有裂缝的影像100幅,每组200幅组成数据集进行测试,采用分类指标统计的方法评测本文算法性能,在正确率、灵敏度、特效度、精度上均达到95%以上,道路裂缝的检测与提取时间约为1 min。结论 该方法不仅可以有效地识别裂缝,同时可以克服了环境中多种因素的干扰,误识别率较低,具有较高的实际应用价值。     

An image processing-based crack detection technique for pressed panel products

Crack detection is an important step in assessing the quality of pressed panel products. This paper presents a fast and non-invasive crack detection technique which involves extracting the outline of the captured object and applying a unique edge line evaluation. This technique is robust against environmental condition changes and only require a low-cost web camera. After capturing an image immediately following the press process, a clear one-pixel edge line is extracted by applying a light control and a series of pre-image processing algorithms, including a valley-emphasis Otsu method and percolation-based shape recognition. Next, the initial detection at low resolution is applied to search for every possible crack using unique edge line and curvature evaluation. Finally, at high resolution, the windowed image of every possible crack is individually analyzed to detect existing cracks using a more specific evaluation process. All of these steps are completed within 0.5 s, thus allowing for the technique to be applied in real-time on a highly automated manufacturing line. To demonstrate the performance of the proposed technique, experiments are conducted on an aluminum plate with different patterns and the pressed panel products. The results show that the proposed technique can detect surface cracks on pressed panels with stable performance as well as high accuracy and efficiency.