Laser-induced damage tests based on a marker-based watershed algorithm with gray control

An effective damage test method based on a marker-based watershed algorithm with gray control(MWGC) is proposed to study the properties of damage induced by near-field laser irradiation for large-aperture laser facilities.Damage tests were performed on fused silica samples and information on the size of damage sites was obtained by this new algorithm,which can effectively suppress the issue of over-segmentation of images resulting from non-uniform illumination in darkfield imaging.Experimental analysis and results show that the lateral damage growth on the exit surface is exponential,and the number of damage sites decreases sharply with damage site size in the damage site distribution statistics.The average damage growth coefficients fitted according to the experimental results for Corning-7980 and Heraeus-Suprasil312 samples at 351 nm are 1.10 ± 0.31 and 0.60 ± 0.09,respectively.

Image Segmentation Using an Improved Watershed Algorithm

As watershed algorithm suffers from over-segmentation problem, this paper presented an efficient method to resolve this problem. First, pre-process of the image using median filter is made to reduce the effect of noise. Second, watershed algorithm is employed to provide initial regions. Third, regions are merged according to the information between the region and boundary. In the merger processing based on the region information, an adaptive threshold of the difference between the neighboring regions is used as the region merge criteria, which is based on the human visual character. In the merger processing on the boundary information, the gradient is used to judge the true boundary of the image to avoid merging the foreground with the background regions. Finally, post-process to the regions using mathematical morphology open and close filter is done to smooth object boundaries. The experimental results show that this method is very efficient.

A Tongue-images Segmentation Method Based on Local Restoration and Watershed Algorithm

Objective: To propose a method to segment tongue- images efficiently, and extract tongue- body accurately and quickly. Methods: Firstly, a kind of color- images' pre- processing technique was used to solve tongue- surface reflection problem. Neighbouring and similar region's information was used to restore the region with tongue- surface reflection problem by replacement. Secondly, the restored image was transformed into a gray one, and then processed by mathematical morphological operation- dilation to get a closed- loop edge. The third technique used was watershed algorithm, which is an usual tool in image segmentation. 'Watershed' function of matlab software was used to complete this algorithm. After that, region- combination technique was used. Through measuring neighbourship and similarity of regions, a non- objective and non- background region was merged into one of its neighbouring regions. This step was repeated until only two regions, objective and background regions, were left in the image. At last, corresponding to the merged image, tongue- body image was got from the original image. Results: 316 images were randomly taken from the image library for experiments, and 299 images were correctly segmented, so, the successful ratio is 94.62%. On the other hand, average time of running this method was about 50 s under whole sampling environment. Conclusion: The method presented in this paper can segment a tongue- body image from its original one effectively, and thus laying a good foundation for the following analysis work.

Image Segmentation Algorithm Based on Improved Watershed Algorithm for Young Leaves of Tea

Aiming at the problem that the traditional watershed image segmentation algorithm is sensitive to noise and prone to ＂over-segmentation＂, an image segmentation method based on improved watershed algorithm is proposed. First, the denoising method is used to denoise the tea image by using the differential equation denoising model The interference of the image on the image segmentation, the protection of the tea image of the edge of the tea information; and then use the watershed algorithm to denoise the tea image after the split. The simulation results show that this method can effectively avoid the influence of noise on image segmentation, and get a good image of ,louug leaves of tea image.

克服Watershed算法过度分割的方法

Water-shed算法是一种形态学的图像分割算法,但由于其对噪声十分敏感,分割结果往往存在过度分割的现象。该文提出一种基于小波分析的Watershed算法,充分利用小波的多分辨率特性有效地解决了Watershed算法的过度分割问题,并大大提高了分割计算速度。

滑动窗口式三维watershed算法实现视频序列分割

视频序列分割是实现视频对象提取、处理和识别的基础,也是基于内容的视频压缩和检索的前提。目前普遍采用的watershed算法是全局涨水算法,用于视频分割时占用的存储和处理空间巨大。针对这一问题本文提出了新的滑动窗口式三维watershed算法(SW3DW),通过时间轴上的局部窗口对视频序列进行观察和分割,而在相邻窗口间采用水源区标号传递和局部流域更新相结合的方法,以较小的存储和处理空间实现了局域和全局处理的一致性。理论分析和实验证明该算法能有效性地实现视频序列的分割,为后续视频对象提取和处理创造了条件。

Ore Image Segmentation Method Based on U-Net and Watershed

Ore image segmentation is a key step in an ore grain size analysis based on image processing.The traditional segmentation methods do not deal with ore textures and shadows in ore images well Those methods often suffer from under-segmentation and over-segmentation.In this article,in order to solve the problem,an ore image segmentation method based on U-Net is proposed.We adjust the structure of U-Net to speed up the processing,and we modify the loss function to enhance the generalization of the model.After the collection of the ore image,we design the annotation standard and train the network with the annotated image.Finally,the marked watershed algorithm is used to segment the adhesion area.The experimental results show that the proposed method has the characteristics of fast speed,strong robustness and high precision.It has great practical value to the actual ore grain statistical task.

Spatiotemporal Modeling of Monthly Precipitation in the Upper Shiyang River Watershed in West Central Gansu,Northwest China

This paper outlines a methodology to estimate monthly precipitation surfaces at 1-kin resolution for the Upper Shiyang River watershed （USRW） in northwest China. Generation of precipitation maps is based on the application of a four-variable genetic algorithm （GA） trained on 10 years of weather and ancillary data, i.e., surface air temperature, relative humidity, Digital Elevation Model-derived estimates of elevation, and time of year collected at 29 weather stations in west-central Gansu and northern Qinghai province. An observed-to-GA predicted data comparison of 10 years of precipitation collected at the 29 weather stations showed that about 84% of the variability in observed values could be explained by the trained GA, including variability in two independent datasets. Point-comparisons of observed and modeled precipitation along an elevation-rainfall gradient demonstrated near-similar spatiotemporal patterns. A precipitation surface for USRW for July, 2005, was developed with the trained GA and input surfaces of surface air temperature and relative humidity generated from Moderate Resolution Imaging Spectroradiometer sensor （MODIS） products of land surface temperature. Spatial tendencies in predicted maximum and minimum values of surface air temperature, relative humidity, and precipitation within a 2-kin radius circle around selected weather stations were in close agreement with the values measured at the weather stations.

基于Watershed和Snake混合模型的图像分割算法

针对传统Snake模型存在着对初始轮廓曲线位置敏感和不能收敛到凹陷区域的问题,提出将改进的分水岭算法应用于优化Snake模型边缘曲线中。同时再利用遗传算法的全局优化特性,来解决Snake模型的局部极小值问题,提升Snake模型的分割精度。实验结果表明,改进的Snake分割算法不仅能有效解决传统Snake分割算法对初始轮廓位置敏感以及不能收敛到凹陷区域的问题,而且具有分割效率高、分割结果更加精确的优点。

IMAGE SEGMENTATION BASED ON MARKOV RANDOM FIELD AND WATERSHED TECHNIQUES

This paper presented a method that incorporates Markov Random Field(MRF), watershed segmentation and merging techniques for performing image segmentation and edge detection tasks. MRF is used to obtain an initial estimate of x regions in the image under process where in MRF model, gray level x , at pixel location i , in an image X , depends on the gray levels of neighboring pixels. The process needs an initial segmented result. An initial segmentation is got based on K means clustering technique and the minimum distance, then the region process in modeled by MRF to obtain an image contains different intensity regions. Starting from this we calculate the gradient values of that image and then employ a watershed technique. When using MRF method it obtains an image that has different intensity regions and has all the edge and region information, then it improves the segmentation result by superimpose closed and an accurate boundary of each region using watershed algorithm. After all pixels of the segmented regions have been processed, a map of primitive region with edges is generated. Finally, a merge process based on averaged mean values is employed. The final segmentation and edge detection result is one closed boundary per actual region in the image.

Automated Watershed Evaluation of Flat Terrain

Catchment area and watershed delineation is a common task in hydrology. The determination of the catchment is still as a challenge considered being crucial key issues particular in flat terrains. Three concepts could be briefly identified through: the manual delineation of a catchment based on a topographic map with contour lines which is a difficult task for flat terrains, by combination of field survey or evaluation of satellite images. The present research is focus on evaluating the possibility to delineate catchments from flat and arid areas by means of DTM avoiding hard techniques like river burning or other manual hydrological DTM corrections. Three GIS packages were used (Arc Hydrotools, TNTmips and RiverTools) within two DEM: the 90 m and 30 m SRTM in addition to the ASTER 30 m, the application sample presented by western Iraq desert—Ubaiydh wadi. A brief review is given how the delineation algorithms have been developed since the 1980’s. Where result shows that automated watershed analysis of flat terrains is cannot be done without manual evaluation and correction either by using several seeding points or river burning technique.

Benefit Evaluation Model of Small Watershed Control Based on Projection Pursuit

A projection pursuit model is presented in this paper for comprehensive evaluation of benefits of small watershed control.By using the model,small watershed control samples with many benefit evaluation indexes can be synthesized projective values with one dimension.The samples can be naturally evaluated according to the projective values.The parameters of the model is optimized by using real coding beased accelerating genetic aglrothm,which overcomes the shortcomings of large computation amount and difficulty of computer programming in traditional projection prusuit methods,and provides a new way for wide applications of projection pursuit technique to different evaluation problems in agricultural systems engineering.

无人机影像GrabCut路面裂缝识别

无人机影像裂缝提取是近年来研究的热点问题之一。针对无人机影像强边缘信息干扰问题,本文提出了一种基于GrabCut算子的裂缝识别方法。该方法首先运用GrabCut算子提取保留裂缝的前景路面,然后运用去噪、边缘检测和双阈值轮廓识别方法探测路面裂缝。这种裂缝识别方法较好地排除了大量伪边缘信息和次生噪声干扰,实现了高分辨率无人机影像裂缝自动识别。试验结果表明,基于GrabCut算子的路面提取方法优于颜色特征提取算法和分水岭算法,适用于复杂场景路面提取,具有较高的普适性;同时,该方法可以快速获取裂缝信息,检测尺度可以人为控制,易于实现多尺度裂缝信息识别。研究结果可应用于路面裂缝定位识别、线性路面设施检测及路面灾害性评估等领域。

基于无人机可见光影像的人工林单木树冠提取研究

以人工林区的无人机可见光影像为基础构建冠层高度模型(canopy height model,CHM),以此作为提取单木树冠的基础数据,避免数字正射影像(digital orthophoto map,DOM)中存在的树冠阴影、光照不均、树冠与草地颜色相似等问题对单木树冠提取的影响,采用标记控制的分水岭算法从CHM中提取单木树冠.首先,采用Otsu算法对CHM中的树冠与非树冠进行分离,利用形态学图像细化算法去除树冠之间的粘连,得到粗略的单木树冠;其次,结合距离变换和H-Maxima算法获取单木位置标记,结合距离变换和分水岭算法获取每个粗略单木树冠的影响区域边界,作为背景标记;为增强CHM中的树冠边缘像素,采用Sobel算法生成梯度图像,利用极小值强加算法将单木位置标记设定为分水岭算法分割的起始点,将背景标记用于对分割过程进行约束;最后,对标记过的梯度图像进行分水岭分割,获得单木树冠.采用人工目视解译的方式获取研究区内正确的单木树冠轮廓作为精度验证数据.通过精度验证,提取的单木位置总体精度为92.79%,提取的单木树冠F-score值为84.62%,提取精度较高.3组对比试验,分别是不进行去除树冠粘连处理、用局部极大值算法代替H-Maxima算法、不使用背景标记,得到3个单木树冠提取结果,通过精度验证,F-score值分别为72.33%、10.16%、77.89%,分别比提出方法降低了12.29%、74.46%、6.73%.通过方法对比和通用性检验,试验结果表明以CHM为研究数据,利用标记控制分水岭算法可有效获取人工林区单木树冠.

基于机器视觉的镍板表面气孔分割算法研究

针对金属镍板表面气孔缺陷检测存在分割目标堆叠和分割线偏移等问题,提出了一种基于机器视觉的镍板表面气孔分割方法,以质心标记计算距离图。该算法采用工业相机构建机器视觉系统,获取镍板表面图像。通过分段线性灰度变换抑制不感兴趣区域,其次对G通道分量进行自适应阈值分割、形态学处理等提取气孔区域,然后图像作差提取出极小值标记,以极小值标记的质心计算距离图并分水岭分割。实验结果表明,该算法气孔正确分割率达94.7%,同时能够获得更准确的气孔分割效果。

基于3D-motif法的刮研表面形貌表征

针对刮研表面微观特性和功能机理研究中缺少表面形貌量化表征方法的问题,采用3D-motif法对刮研表面形貌进行表征。使用LI-3型接触式三维表面形貌测量仪对刮研表面进行测量,并利用三维点云数据生成刮研表面的二维灰度图像。然后,根据3D-motif法中集水盆地的定义,采用分水岭算法对刮研表面的灰度图像进行motif分割与合并。以不同精度等级的刮研表面的整体纹理区域motif分割结果为对象,定义了特征显著度并提取计算了刮研表面在2种不同面积尺度(25 mm^(2)和0.25 mm^(2))上的深度、面积、方向角、各向异性率、扁平系数和特征显著度等6项motif参数。结合部分motif参数的分布情况和motif数量的变化趋势,从形貌特征的尺寸和形态两个方面对刮研表面进行了表征和分析,实现了以较少参数完整表征刮研表面的三维形貌。研究结果可为后续刮研表面微观特性的深入分析提供理论基础。

基于GEE的洱海流域土地利用/覆被分类算法对比研究

快速准确地进行复杂高原山区的土地覆被长时序自动分类,可为国土规划、资源利用提供依据。利用GEE云平台,选取Landsat影像地表反射率、植被指数、水体指数、DEM 4种空间数据集作为土地覆被分类的基础和辅助数据,分别运用CART、RF和SVM 3种分类算法,实现洱海流域土地覆被信息的自动提取和精度对比。结果表明:(1)3种分类算法中,RF的总体分类精度最高,SVM的总体精度最低;RF是洱海流域LULC的最适宜分类算法。(2)采用光谱指数、地形特征等辅助数据集会进一步提高解译精度,而样本点的选取是最主要的影响因素。(3)Erhai_RF能够达到较高的精度,同时更加突出细节特征,在局部实际分类精度上会更高。研究结果可为洱海流域长时序土地覆被数据产品智能快速提取以及最优分类算法筛选提供方法和技术支撑。

基于GMM和分水岭算法的润滑油粘连磨粒的在线检测方法

润滑油中磨粒的数量、尺寸特征是磨损程度判断的主要依据,由于运动状态下的磨粒极易发生粘连,难以实现磨粒的在线监测及特征提取。为提高粘连磨粒目标检测精度,提出了一种基于融合GMM(Gaussian Mixture Modelling,GMM)和分水岭算法(Watershed Algorithm)的磨粒目标检测算法。首先,采用基于GMM的背景差分算法初步提取出运动磨粒目标,然后再利用分水岭算法对所提取目标进行分割,完成粘连磨粒目标的精确检测。实验结果表明,相比于传统算法,GMM-Watershed算法能够有效提高粘连磨粒的目标检测精度,准确率达98.86%,为提高油液磨粒在线特征提取的准确率提供了有效的方法。

基于U-Net网络与分水岭相结合的脑肿瘤分割

有效的MRI脑肿瘤图像分割能够为医生对患者的诊断和治疗提供可靠的依据。虽然卷积神经网络在医学图像分割的领域取得了显著的进展,但脑部结构过于复杂,误分割率较高,MRI脑肿瘤图像分割仍存在许多不足。U-Net网络的全对称结构能够使其只需少量训练即可提取足够的特征,但由于U-Net网络每次卷积图像都会小一圈,导致上采样和下采样所还原的像素尺寸不一样,无法对肿瘤边缘进行准确地分割。为解决上述问题,提出了一种基于U-Net网络与分水岭相结合的脑肿瘤分割算法。利用U-Net网络模型中的跳跃连接结合压缩路径和扩展路径的特征图,得到对感兴趣区域的初分割,然后通过添加基于重建的开闭操作的分水岭算法优化初分割图边界,得到最终的分割结果。实验结果表明,该方法在准确率Acc、特异性Sp、灵敏度Sn和精度PPV上分别可达到0.896、0.988 8、0.905 9和0.980 1,能够有效地分割出病变区域,具有明显的研究价值。

基于改进PSO-KMeans煤炭异物筛选算法研究

采煤过程中异物自动识别和分拣是实现矿业信息化的关键技术之一。传统双能X射线系统根据R值算法可有效识别出煤炭中混杂的钢筋与胶皮,却难以识别与煤炭组成成分相似的木质杂质。针对这一问题,提出基于L_(0)范数最小化与改进PSO-KMeans的木质杂质筛选算法,借助L_(0)范数最小化算法平滑图像,去除煤灰干扰,利用改进PSO-KMeans聚类算法与基于距离变换的分水岭算法实现图像分割,根据离心率与矩形度进行木质杂质识别,并通过仿真实验验证方法的可行性。经验证此方法能有效筛选出煤炭中混杂的木质杂质。