WO2023279556A1 - 一种骨料粒径抽样监测方法 - Google Patents

一种骨料粒径抽样监测方法 Download PDF

Info

Publication number
WO2023279556A1
WO2023279556A1 PCT/CN2021/122223 CN2021122223W WO2023279556A1 WO 2023279556 A1 WO2023279556 A1 WO 2023279556A1 CN 2021122223 W CN2021122223 W CN 2021122223W WO 2023279556 A1 WO2023279556 A1 WO 2023279556A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
aggregate
particle size
value
belt
Prior art date
Application number
PCT/CN2021/122223
Other languages
English (en)
French (fr)
Inventor
魏朝明
庄杰彬
侯贵鹏
黄骁民
温金开
张宝裕
杨建红
Original Assignee
福建南方路面机械股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 福建南方路面机械股份有限公司 filed Critical 福建南方路面机械股份有限公司
Publication of WO2023279556A1 publication Critical patent/WO2023279556A1/zh

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N15/00—Investigating characteristics of particles; Investigating permeability, pore-volume or surface-area of porous materials
    • G01N15/02—Investigating particle size or size distribution
    • G01N15/0205—Investigating particle size or size distribution by optical means
    • G01N15/0227—Investigating particle size or size distribution by optical means using imaging; using holography
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/70—Denoising; Smoothing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/136—Segmentation; Edge detection involving thresholding
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes
    • G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20024—Filtering details
    • G06T2207/20028—Bilateral filtering
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20024—Filtering details
    • G06T2207/20032—Median filtering
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20036—Morphological image processing
    • G06T2207/20041—Distance transform
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20112—Image segmentation details
    • G06T2207/20152—Watershed segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30108—Industrial image inspection
    • G06T2207/30132—Masonry; Concrete

Definitions

  • the invention relates to the technical field of aggregate monitoring methods, in particular to an aggregate particle size sampling monitoring method.
  • Aggregate is the main material used in asphalt mixture and cement concrete, and the accuracy of aggregate grading has a great influence on the quality of the final product.
  • the present invention provides a low calculation amount, high accuracy, real-time monitoring and low time delay aggregate particle size sampling monitoring method.
  • a method for sampling monitoring of aggregate particle size is characterized in that it comprises the following steps:
  • the distance sensor measures the distance towards the belt, and compares the measured data with the corresponding preset data in the database to determine whether the belt is running and whether there is aggregate on the belt; if the belt is running and carries aggregate, then proceed to step b;
  • the camera shoots images, and the camera shoots images towards the belt to obtain a square image, and each edge of the square image is within the range of the belt;
  • Image pre-analysis extract the edge pixels of the square image, select two horizontal lines composed of several rows of pixels on the upper and lower edges of the image, select two vertical lines composed of several columns of pixels on the left and right edges of the image, and calculate the pixels on the four lines respectively Gray value, respectively calculate the difference value of the pixel gray value on the four lines, respectively calculate the variance of the pixel gray value on each line; when the absolute value of the difference value of the pixel gray value on each line is satisfied When the sum is greater than the corresponding preset value in the database, and the variance of the absolute value of the difference value of the gray level of each pixel on each line is greater than the corresponding preset value in the database, it is judged as a valid image, otherwise it is judged that the image is incomplete and invalid image;
  • Image processing processing the effective image to obtain the particle information of each aggregate on the map
  • the belt running detection includes: the distance sensor measures the distance towards the belt, obtains the material level height information of the aggregate on the belt through the measurement data, and the distance sensor conducts continuous measurement through a set frequency, and the collection interval is 50ms-200ms. Take the distance measurement results within 1s to calculate the average value and range, and judge whether the average value and range meet the preset values in the database. If they are satisfied for 2 consecutive seconds, it is determined that the belt is running and there is aggregate on the belt.
  • the image processing includes: performing gray level homogenization processing on the effective image, dividing the image evenly into several sub-sections, counting the gray level distribution of each sub-section, and selecting the gray level with moderate gray level as the
  • the target gray level strengthens the gray level of each sub-interval; the noise points and singular points of the image are removed by median filtering and bilateral filtering; Redefine the area whose gray value is greater than the set threshold; extract the edge of the obtained image, and obtain a new gray image through distance transformation, perform watershed calculation and segmentation on the image after distance transformation, and divide the new gray image
  • the high-resolution image is superimposed on the original image and the black background area is subtracted to obtain a segmented image corresponding to the particle position of the aggregate.
  • the image processing also includes: by calculating the equivalent elliptical Feret diameter of each aggregate particle as the particle size of the identified aggregate, counting the particle size and area of each aggregate in each image, and calculating The product represents the aggregate particle volume, that is, the particle information of aggregate particle size and volume is obtained.
  • the comparison judgment includes: dividing the aggregate into several sub-particle size intervals according to the particle size, and calculating the ratio of the aggregate volume in each sub-particle size interval to the total volume to obtain the actual aggregate grading curve.
  • the comparative judgment also includes: calculating the difference between the volume ratio of each sub-particle size interval of the actual aggregate gradation curve and the volume ratio of each sub-particle size interval corresponding to the preset aggregate gradation curve, and Calculate the sum of the absolute values of the differences, the difference and the sum of the absolute values of the differences are the offset degree value, judge whether the maximum value of the difference and the sum of the absolute values are less than the preset value in the database, and the two If all of them meet the conditions, it means that the monitored aggregate particle size is qualified, otherwise, it will be judged as unqualified and an alarm message will be sent.
  • the aggregate particle size sampling monitoring method provided by the present invention performs online real-time sampling monitoring of the aggregates transported on the belt, which is compared with the detection of stacked materials.
  • the method greatly reduces the time delay, and when there is a problem with the aggregate grading, it can be found in time and dealt with accordingly, ensuring the quality of the final product.
  • the distance sensor is used to monitor the material level of the belt, and the judgment of whether the belt is running and whether the belt is carrying aggregate is realized with a very low calculation amount, and the efficiency is high.
  • the camera shoots a square image on the belt, and the image pre-analysis can identify whether the image is a valid image by simply judging the difference between the gray value of the edge of the square image, eliminating the interference of invalid images on the monitoring results, and image pre-analysis It greatly reduces the calculation amount of subsequent image processing and improves the accuracy.
  • a kind of aggregate particle size sampling monitoring method comprises the steps:
  • the distance sensor measures the distance towards the belt, and obtains the material level height information of the aggregate on the belt through the measurement data, and the distance sensor conducts continuous measurement through a set frequency, and the collection interval is 50ms-200ms, preferably 100ms Carry out a measurement, take the distance measurement results within 1s to calculate the average value and range, and judge whether the average value and range meet the preset values in the database. If they are satisfied for 2 consecutive seconds, it is determined that the belt is running and there is aggregate on the belt , if the belt is running and carrying aggregate, proceed to step b;
  • the camera shoots images, and the camera shoots images towards the belt to obtain a square image, and each edge of the square image is within the range of the belt;
  • Image pre-analysis extract the edge pixels of the square image, select two horizontal lines composed of several rows of pixels on the upper and lower edges of the image, preferably four rows of pixels, and select two vertical lines composed of several columns of pixels on the left and right edges of the image, preferably four rows Pixel, respectively calculate the gray value of each pixel on the four lines, respectively calculate the difference value of the gray value of the pixel point on the four lines, and calculate the variance of the gray value of the pixel point on each line; when each pixel on each line is satisfied When the sum of the absolute value of the differential value of the point gray is greater than the corresponding preset value in the database, and the variance of the absolute value of the differential value of the gray value of each pixel on each line is greater than the corresponding preset value in the database, it is judged to be valid image, otherwise it is judged that the image is incomplete and invalid;
  • image pre-analysis is: the aggregate on the belt is not uniform at the beginning stage of aggregate conveying, the tailing stage of conveying is about to end, midway pause and restart stage, which is manifested by the lack of aggregate in a large part of the position. If the The results of monitoring and calculating the output at these stages will have large deviations, and wrong results will be output and false alarms will be generated. Therefore, image pre-analysis is required to identify images of these invalid stages to eliminate them; since the color of the belt is relatively consistent with the aggregate, If the belt background is captured in the image, the gray scale of each pixel has a small change and the difference is small. The gray scale value on the straight line near the four edges of the square image can be used to judge whether it is With belt background, images with belt background are excluded.
  • the length of the sub-particle size interval is 0.1 mm
  • the total measurement range is 0-50 mm, which is divided into 500 sub-particle size intervals.
  • Each aggregate (within a particle size of 50 mm) has a corresponding volume sum, and the ratio of the volume sum of each sub-particle size range to the total volume sum is the volume ratio of each sub-particle size range.
  • the information changes measured by the 2D image method represent the real aggregate particle size changes. Compare the statistical results of the measurement results for a period of time with the preset aggregate grading curve of the pre-collected standard sample, and calculate the difference in the volume ratio of each sub-particle size interval between the two. If the measured aggregate size is different from the standard sample There is no big difference in size, the aggregate gradation curves of the two should be relatively close or coincident, and the obtained difference is also within the normal range. If the particle size difference between the two is large, the calculated actual aggregate grading curve will be far away from the preset aggregate grading curve. By calculating the difference between the two, the specific difference range can be obtained.
  • each preset threshold for offset judgment is set in advance in the software, and can also be changed according to actual needs during the monitoring process.
  • Ultra-particle size threshold Set the ultra-particle size threshold to a fixed particle size value. When the volume ratio exceeding this value exceeds the set range, an ultra-particle size alarm will be prompted. For example, when monitoring aggregates of 20-30mm, set the oversize threshold to 40mm, and set the exceeding ratio threshold to 20%. Super particle size alarm prompt.
  • a method for sampling and monitoring aggregate particle size according to the invention through the steps of belt running detection, camera shooting image, image pre-analysis, image processing, comparison and judgment, aggregate particle size sampling monitoring, can be widely used in asphalt mixture, cement
  • the production of concrete products has good industrial applicability.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Chemical & Material Sciences (AREA)
  • Dispersion Chemistry (AREA)
  • Quality & Reliability (AREA)
  • Geometry (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Biochemistry (AREA)
  • General Health & Medical Sciences (AREA)
  • Immunology (AREA)
  • Pathology (AREA)
  • Image Processing (AREA)
  • Length Measuring Devices By Optical Means (AREA)

Abstract

一种骨料粒径抽样监测方法，包括皮带运行检测、摄像机拍摄图像、图像预分析、图像处理、比对判断步骤。通过图像预分析，选取图像上下左右边缘的四条线，分别计算四条线上的各像素灰度值,分别计算四条线上像素点灰度值的差分值，分别计算每条线上像素点灰度值的方差；当满足每条线上各像素点灰度的差分值的绝对值求和大于数据库内对应的预设值，并且每条线上各像素点灰度的差分值的绝对值的方差大于数据库内对应的预设值时判断为有效图像。图像预分析通过对方形图像边缘灰度值的差值进行简易判断，排除了无效图像对监测结果的干扰，降低了后续图像处理的计算量，提高准确度。

Description

一种骨料粒径抽样监测方法 技术领域
本发明涉及骨料监测方法技术领域，特别涉及一种骨料粒径抽样监测方法。
背景技术
骨料是沥青混合料与水泥混凝土的主要用料，骨料级配的准确性对最终成品料的质量有着较大影响。目前市场上使用的水泥混凝土(或沥青混合料)生产设备，在运行过程中，对于骨料粒径的判断一般是在料堆上进行取样筛分判断，设备上没有有效的对骨料粒径是否合格进行监测的装置，实际上的料堆骨料量大，需要消耗较多的人力，也很难做到具有代表性的检测，若料堆中存在部分粒径不合格的骨料，通过这种检测方式很难发现，容易导致设备生产的成品不合格，也有可能导致设备的损坏。
发明内容
为克服现有技术中的不足，本发明提供一种低计算量，高准确度，实时监测且低延时性的骨料粒径抽样监测方法。
为实现上述目的，本发明采用的技术方案如下：一种骨料粒径抽样监测方法，其特征在于包括如下步骤：
a.皮带运行检测，距离传感器朝向皮带进行间距测量，通过测量 数据与数据库内对应的预设数据进行比对判断皮带是否运行及皮带上是否运载有骨料；若皮带运行且运载有骨料则进行步骤b；
b.摄像机拍摄图像，摄像机朝向皮带进行图像拍摄，获取方形图像，所述方形图像的各边缘均处于皮带范围内；
c.图像预分析，对方形图像边缘像素进行提取，选取图像上下边缘若干行像素组成的两条水平线，选取图像左右边缘若干列像素组成的两条竖直线，分别计算四条线上的各像素灰度值,分别计算四条线上像素点灰度值的差分值，分别计算每条线上像素点灰度值的方差；当满足每条线上各像素点灰度的差分值的绝对值求和大于数据库内对应的预设值，并且每条线上各像素点灰度的差分值的绝对值的方差大于数据库内对应的预设值时判断为有效图像，否则判断为图像不完整为无效图像；
d.图像处理，对有效图像进行处理，获得图上各骨料的颗粒信息；
e.比对判断，对一段时间内所有的有效图像的各骨料的颗粒信息进行整合，获得实际骨料级配曲线，将实际骨料级配曲线与数据库中预设骨料级配曲线进行比对计算，获得偏移度数值，通过偏移度数值判断骨料是否合格，若不合格则发送报警信息提示。
进一步的，皮带运行检测包括：距离传感器朝向皮带进行间距测量，通过测量数据获取皮带上骨料的料位高度信息，距离传感器通过一设定频率进行连续化测量，采集间隔时间为50ms-200ms，取1s内的间距测量结果进行平均值及极差计算，判断平均值与极差是否满足数据库内预设值，若连续2s均满足，则判定皮带运行且皮带上运载 有骨料。
进一步的，图像处理包括：对有效图像进行灰度均匀化处理，将图像均匀分为若干个子分区，统计出每个子分区的灰度分布，根据各个分区的差异，选择灰度适中的灰度作为目标灰度对各个子区间的灰度进行加强；通过中值滤波与双边滤波方式去除图像的噪声点以及奇异点；进行灰度重构，利用过滤灰度极大值点的方式将图像中的灰度值大于设定阈值的区域进行重新定义；将得到的图像进行边缘提取，并通过距离变换得到一幅新的灰度图像，对距离变换后的图像进行分水岭计算并分割，将新的灰度图像叠加到原图并扣除黑色背景区域，得到对应骨料的颗粒位置的分割图像。
进一步的，图像处理还包括：通过计算各个骨料颗粒的等效椭圆Feret径作为识别骨料的粒径，统计每一张图像的每个骨料的粒径和面积，通过面积与粒径的乘积表示骨料颗粒体积，即获得骨料粒径及体积的颗粒信息。
进一步的，比对判断包括：根据骨料粒径分为若干子粒径区间，计算各子粒径区间内的骨料的体积占总体积的比例，得到实际骨料级配曲线。
进一步的，对比判断还包括：将实际骨料级配曲线的各子粒径区间的体积占比，与预设骨料级配曲线对应各子粒径区间的体积占比进行差值计算，且计算各差值的绝对值之和，差值、差值的绝对值之和为所述偏移度数值，判断差值的最大值和绝对值之和是否均小于数据库中的预设值，两者均满足条件即表示监测骨料粒径合格，否则，判 定不合格并发送报警信息提示。
由上述对本发明的描述可知，与现有技术相比，本发明提供的一种骨料粒径抽样监测方法，对皮带上运送的骨料进行在线实时抽样监测，相比于对堆料进行检测的方式极大降低了时延性，当出现骨料级配出问题后可及时发现以进行相应处理，保证了最终产品的质量。通过距离传感器对皮带进行料位高度的监测，以极低的计算量实现了对皮带是否运行及皮带是否运载骨料的判断，效率高。摄像机在皮带上拍摄方形图像，图像预分析通过对方形图像边缘灰度值的差值的进行简易判断，即可识别出图像是否为有效图像，排除了无效图像对监测结果的干扰，图像预分析极大降低了后续图像处理的计算量，提高准确度。
具体实施方式
以下通过具体实施方式对本发明作进一步的描述。
一种骨料粒径抽样监测方法，包括如下步骤：
a.皮带运行检测，距离传感器朝向皮带进行间距测量，通过测量数据获取皮带上骨料的料位高度信息，距离传感器通过一设定频率进行连续化测量，采集间隔时间为50ms-200ms，优选100ms进行一次测量，取1s内的间距测量结果进行平均值及极差计算，判断平均值与极差是否满足数据库内预设值，若连续2s均满足，则判定皮带运行且皮带上运载有骨料，若皮带运行且运载有骨料则进行步骤b；
b.摄像机拍摄图像，摄像机朝向皮带进行图像拍摄，获取方形图像，所述方形图像的各边缘均处于皮带范围内；
c.图像预分析，对方形图像边缘像素进行提取，选取图像上下边缘若干行像素组成的两条水平线，优选四行像素，选取图像左右边缘若干列像素组成的两条竖直线，优选四行像素，分别计算四条线上的各像素灰度值,分别计算四条线上像素点灰度值的差分值，分别计算每条线上像素点灰度值的方差；当满足每条线上各像素点灰度的差分值的绝对值求和大于数据库内对应的预设值，并且每条线上各像素点灰度的差分值的绝对值的方差大于数据库内对应的预设值时判断为有效图像，否则判断为图像不完整为无效图像；
d.图像处理，对有效图像进行灰度均匀化处理，将图像均匀分为若干个子分区，统计出每个子分区的灰度分布，根据各个分区的差异，选择灰度适中的灰度作为目标灰度对各个子区间的灰度进行加强；通过中值滤波与双边滤波方式去除图像的噪声点以及奇异点；进行灰度重构，利用过滤灰度极大值点的方式将图像中的灰度值大于设定阈值的区域进行重新定义；将得到的图像进行边缘提取，并通过距离变换得到一幅新的灰度图像，对距离变换后的图像进行分水岭计算并分割，将新的灰度图像叠加到原图并扣除黑色背景区域，得到对应骨料的颗粒位置的分割图像；通过计算各个骨料颗粒的等效椭圆Feret径作为识别骨料的粒径，统计每一张图像的每个骨料的粒径和面积，通过面积与粒径的乘积表示骨料颗粒体积，即获得骨料粒径及体积的颗粒信息；
e.比对判断，对一段时间内所有的有效图像的各骨料的颗粒信息进行整合，根据骨料粒径分为若干子粒径区间，计算各子粒径区间内的骨料的体积占总体积的比例，得到实际骨料级配曲线，将实际骨料级配曲线的各子粒径区间的体积占比，与预设骨料级配曲线对应各子粒径区间的体积占比进行差值计算，且计算各差值的绝对值之和，差值、差值的绝对值之和为偏移度数值，判断差值的最大值和绝对值之和是否均小于数据库中的预设值，两者均满足条件即表示监测骨料粒径合格，否则，判定不合格并发送报警信息提示。
图像预分析的原理为：骨料在输送开始阶段，输送即将结束的尾料阶段、中途暂停及重启阶段，皮带上的骨料并不均匀，表现为较大部分的位置骨料缺失，若将这些阶段进行监测并计算输出的结果将会有较大的偏差，输出错误结果而产生误报警，因此需要图像预分析识别出这些无效阶段的图像以排除；由于皮带的颜色相对骨料较为一致，图像中若拍摄到皮带背景，则各像素点灰度变化幅度较小，差值较小，采用取方形图像四条边缘附近的直线上的灰度值，根据其灰度波动的幅度即可判断是否有皮带背景，将存在皮带背景的图像排除。
本实施例中，子粒径区间长度为0.1mm，总测量范围在0-50mm，一共分为500个子粒径区间。每一个骨料(粒径50mm以内)均有对应的体积和，各个子粒径区间的体积和与总体积和的比值为各个子粒径区间的体积占比。
通过大量的实验数据可以用2D图像法测量的信息变化表示真实的骨料颗粒尺寸的变化情况。将一段时间的测量结果统计结果与预先 采集的标准样本的预设骨料级配曲线进行对比，计算两者各子粒径区间体积占比的差值，如果测量的骨料尺寸与标准的样本尺寸无较大差异，两者的骨料级配曲线应该较为接近或者重合，得到的差值也在正常值范围内。如果两者粒径差异较大，则计算出的实际骨料级配曲线会与预设骨料级配曲线距离较远，通过计算两者差值，可以得到具体的相差幅度。
当实际骨料级配曲线与预设骨料级配曲线相差较大时，两者差值的绝对值和达到一定数值后，或者某一段差值大于一定值时，提示粒径偏移报警。本实施例中，偏移判断的各个预设阈值在软件中提前设定，在监测过程中也可根据实际需要进行更改。
将超粒径阈值设为固定的粒径值，当超过此值的体积占比超出设定范围时，即提示超粒径报警。比如，在监测20-30mm的骨料时，设定超粒径阈值为40mm，超过比例阈值设为20％，即当监测的骨料粒径超过40mm的颗粒体积占比大于20％时，开始超粒径报警提示。
上述仅为本发明的一种具体实施方式，但本发明的设计构思并不局限于此，凡利用此构思对本发明进行非实质性的改动，均应属于侵犯本发明保护范围的行为。
工业使用性
本发明一种骨料粒径抽样监测方法，通过皮带运行检测、摄像机拍摄图像、图像预分析、图像处理、比对判断步骤，对骨料粒径抽样监测，可广泛用于沥青混合料、水泥混凝土类产品的生产，具有良好 的工业使用性。

Claims (8)

  1. 一种骨料粒径抽样监测方法，其特征在于包括如下步骤：
    a.皮带运行检测，距离传感器朝向皮带进行间距测量，获取皮带上骨料的料位高度信息，通过测量数据与数据库内对应的预设数据进行比对判断皮带是否运行及皮带上是否运载有骨料；若皮带运行且运载有骨料则进行步骤b；
    b.摄像机拍摄图像，摄像机朝向皮带进行图像拍摄，获取方形图像，所述方形图像的各边缘均处于皮带范围内；
    c.图像预分析，对方形图像边缘像素进行提取，选取图像上下边缘若干行像素组成的两条水平线，选取图像左右边缘若干列像素组成的两条竖直线，分别计算四条线上的各像素灰度值,分别计算四条线上像素点灰度值的差分值，分别计算每条线上像素点灰度值的方差；当满足每条线上各像素点灰度的差分值的绝对值求和大于数据库内对应的预设值，并且每条线上各像素点灰度的差分值的绝对值的方差大于数据库内对应的预设值时判断为有效图像，否则判断为图像不完整为无效图像；
    d.图像处理，对有效图像进行处理，获得图上各骨料的颗粒信息；
    e.比对判断，对一段时间内所有的有效图像的各骨料的颗粒信息进行整合，获得实际骨料级配曲线，将实际骨料级配曲线与数据库中预设骨料级配曲线进行比对计算，获得偏移度数值，通过偏移度数值判断骨料是否合格，若不合格则发送报警信息提示。
  2. 根据权利要求1所述一种骨料粒径抽样监测方法，其特征在 于：所述皮带运行检测步骤，其距离传感器通过一设定频率进行连续化测量，采集间隔时间为50ms-200ms，取1s内的间距测量结果进行平均值及极差计算，判断平均值与极差是否满足数据库内预设值，若连续2s均满足，则判定皮带运行且皮带上运载有骨料。
  3. 根据权利要求1所述一种骨料粒径抽样监测方法，其特征在于：所述图像处理步骤包括：对有效图像进行灰度均匀化处理，将图像均匀分为若干个子分区，统计出每个子分区的灰度分布，根据各个分区的差异，选择灰度适中的灰度作为目标灰度对各个子区间的灰度进行加强；通过中值滤波与双边滤波方式去除图像的噪声点以及奇异点；进行灰度重构，利用过滤灰度极大值点的方式将图像中的灰度值大于设定阈值的区域进行重新定义；将得到的图像进行边缘提取，并通过距离变换得到一幅新的灰度图像，对距离变换后的图像进行分水岭计算并分割，将新的灰度图像叠加到原图并扣除黑色背景区域，得到对应骨料的颗粒位置的分割图像。
  4. 根据权利要求3所述一种骨料粒径抽样监测方法，其特征在于：所述图像处理还步骤包括：通过计算各个骨料颗粒的等效椭圆Feret径作为识别骨料的粒径，统计每一张图像的每个骨料的粒径和面积，通过面积与粒径的乘积表示骨料颗粒体积，即获得骨料粒径及体积的颗粒信息。
  5. 根据权利要求4所述一种骨料粒径抽样监测方法，其特征在于：所述比对判断步骤包括：根据骨料粒径分为若干子粒径区间，计算各子粒径区间内的骨料的体积占总体积的比例，得到实际骨料级配 曲线。
  6. 根据权利要求5所述一种骨料粒径抽样监测方法，其特征在于：所述对比判断步骤还包括：将实际骨料级配曲线的各子粒径区间的体积占比，与预设骨料级配曲线对应各子粒径区间的体积占比进行差值计算，且计算各差值的绝对值之和，差值、差值的绝对值之和为所述偏移度数值，判断差值的最大值和绝对值之和是否均小于数据库中的预设值，两者均满足条件即表示监测骨料粒径合格，否则，判定不合格并发送报警信息提示。
  7. 根据权利要求6所述一种骨料粒径抽样监测方法，其特征在于：所述骨料粒径的子粒径区间长度为0.1mm，总测量范围在0-50mm，一共分为500个子粒径区间；每一个骨料均有对应的体积和，各个子粒径区间的体积和与总体积和的比值为各个子粒径区间的体积占比。
  8. 根据权利要求6所述一种骨料粒径抽样监测方法，其特征在于：所述对比判断步骤，监测20-30mm的骨料时，设定超粒径阈值为40mm，超过比例阈值设为20％，当监测的骨料粒径超过40mm的颗粒体积占比大于20％时，开始超粒径报警提示。
PCT/CN2021/122223 2021-07-09 2021-09-30 一种骨料粒径抽样监测方法 WO2023279556A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110780443.7A CN113674201B (zh) 2021-07-09 2021-07-09 一种骨料粒径抽样监测方法
CN202110780443.7 2021-07-09

Publications (1)

Publication Number Publication Date
WO2023279556A1 true WO2023279556A1 (zh) 2023-01-12

Family

ID=78539014

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/122223 WO2023279556A1 (zh) 2021-07-09 2021-09-30 一种骨料粒径抽样监测方法

Country Status (2)

Country Link
CN (1) CN113674201B (zh)
WO (1) WO2023279556A1 (zh)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115861301A (zh) * 2023-02-16 2023-03-28 山东百成新材料科技股份有限公司 用于改性沥青生产的多物料均匀性视觉评价方法
CN116129364A (zh) * 2023-04-17 2023-05-16 山东山矿机械有限公司 一种皮带集控系统
CN116645363A (zh) * 2023-07-17 2023-08-25 山东富鹏生物科技有限公司 一种基于视觉的淀粉生产质量实时检测方法
CN116681695A (zh) * 2023-07-27 2023-09-01 山东阁林板建材科技有限公司 一种抗变形模板端面的质量检测方法
CN117711127A (zh) * 2023-11-08 2024-03-15 金舟消防工程（北京）股份有限公司 一种消防安全监管方法及系统
CN117772392A (zh) * 2024-02-26 2024-03-29 湖南华菱湘潭钢铁有限公司 一种烧结燃料粒度全智能控制方法
CN117862045A (zh) * 2023-12-22 2024-04-12 山东山田新材科研有限公司 一种超细颗粒连续分选的控制系统
CN118470019A (zh) * 2024-07-12 2024-08-09 济宁职业技术学院 基于图像数据的脱水蔬菜质量快速检测方法

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115797472B (zh) * 2023-02-03 2023-05-12 山东百成新材料科技股份有限公司 一种改性沥青生产用智能筛分方法
CN118777142B (zh) * 2024-09-11 2024-12-06 北京名道恒通信息技术有限公司 一种催化裂化过程中的催化剂测量方法及测量系统

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030156739A1 (en) * 2002-02-15 2003-08-21 Inco Limited Rock fragmentation analysis system
CN107220946A (zh) * 2017-05-18 2017-09-29 长安大学 一种岩石运输带上不良块度图像的实时剔除方法
CN108711149A (zh) * 2018-05-16 2018-10-26 郑州大学 基于图像处理的矿岩粒度检测方法
CN109060813A (zh) * 2018-06-05 2018-12-21 中国矿业大学(北京) 基于图像优化分类的土石料级配自动检测方法
CN110969608A (zh) * 2019-11-29 2020-04-07 华侨大学 一种基于图像法的机制砂级配修正系统
CN111047555A (zh) * 2019-11-13 2020-04-21 鞍钢集团矿业有限公司 基于图像处理技术的矿石图像粒度检测算法
CN111968173A (zh) * 2020-08-14 2020-11-20 湖南长天自控工程有限公司 一种混合料粒度分析方法及系统
CN112017164A (zh) * 2020-08-18 2020-12-01 中国水利水电科学研究院 基于深度阈值卷积模型的土石料级配检测方法

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2008268051A (ja) * 2007-04-23 2008-11-06 Zenkoku Nama Concrete Kogyo Kumiai Rengokai 生コンクリート骨材粒度測定方法および生コンクリート骨材粒度測定システム
CN107621435A (zh) * 2017-10-16 2018-01-23 华侨大学 一种骨料在线监测装置
CN111105386B (zh) * 2019-03-28 2023-04-18 烟台大学 一种基于移动设备的粗骨料质量图像处理及分析方法
CN110118775B (zh) * 2019-05-10 2021-11-19 重庆交通大学 厂拌水泥稳定碎石骨料组成快速检测方法

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030156739A1 (en) * 2002-02-15 2003-08-21 Inco Limited Rock fragmentation analysis system
CN107220946A (zh) * 2017-05-18 2017-09-29 长安大学 一种岩石运输带上不良块度图像的实时剔除方法
CN108711149A (zh) * 2018-05-16 2018-10-26 郑州大学 基于图像处理的矿岩粒度检测方法
CN109060813A (zh) * 2018-06-05 2018-12-21 中国矿业大学(北京) 基于图像优化分类的土石料级配自动检测方法
CN111047555A (zh) * 2019-11-13 2020-04-21 鞍钢集团矿业有限公司 基于图像处理技术的矿石图像粒度检测算法
CN110969608A (zh) * 2019-11-29 2020-04-07 华侨大学 一种基于图像法的机制砂级配修正系统
CN111968173A (zh) * 2020-08-14 2020-11-20 湖南长天自控工程有限公司 一种混合料粒度分析方法及系统
CN112017164A (zh) * 2020-08-18 2020-12-01 中国水利水电科学研究院 基于深度阈值卷积模型的土石料级配检测方法

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115861301A (zh) * 2023-02-16 2023-03-28 山东百成新材料科技股份有限公司 用于改性沥青生产的多物料均匀性视觉评价方法
CN116129364A (zh) * 2023-04-17 2023-05-16 山东山矿机械有限公司 一种皮带集控系统
CN116645363A (zh) * 2023-07-17 2023-08-25 山东富鹏生物科技有限公司 一种基于视觉的淀粉生产质量实时检测方法
CN116645363B (zh) * 2023-07-17 2023-10-13 山东富鹏生物科技有限公司 一种基于视觉的淀粉生产质量实时检测方法
CN116681695A (zh) * 2023-07-27 2023-09-01 山东阁林板建材科技有限公司 一种抗变形模板端面的质量检测方法
CN116681695B (zh) * 2023-07-27 2023-12-01 山东阁林板建材科技有限公司 一种抗变形模板端面的质量检测方法
CN117711127A (zh) * 2023-11-08 2024-03-15 金舟消防工程（北京）股份有限公司 一种消防安全监管方法及系统
CN117862045A (zh) * 2023-12-22 2024-04-12 山东山田新材科研有限公司 一种超细颗粒连续分选的控制系统
CN117772392A (zh) * 2024-02-26 2024-03-29 湖南华菱湘潭钢铁有限公司 一种烧结燃料粒度全智能控制方法
CN117772392B (zh) * 2024-02-26 2024-05-10 湖南华菱湘潭钢铁有限公司 一种烧结燃料粒度全智能控制方法
CN118470019A (zh) * 2024-07-12 2024-08-09 济宁职业技术学院 基于图像数据的脱水蔬菜质量快速检测方法

Also Published As

Publication number Publication date
CN113674201B (zh) 2024-07-26
CN113674201A (zh) 2021-11-19

Similar Documents

Publication Publication Date Title
WO2023279556A1 (zh) 一种骨料粒径抽样监测方法
CN110118775B (zh) 厂拌水泥稳定碎石骨料组成快速检测方法
CN117635565B (zh) 一种基于图像识别的半导体表面缺陷检测系统
CN103499585B (zh) 基于机器视觉的非连续性锂电池薄膜缺陷检测方法及其装置
CN109598715B (zh) 基于机器视觉的物料粒度在线检测方法
CN103954542B (zh) 基于无参考图像清晰度评价的pm2.5浓度检测仪
CN104268505A (zh) 基于机器视觉的布匹疵点自动检测识别装置及方法
EP3178719A1 (en) Train type identification method and system, and security inspection method and system
CN114494210A (zh) 基于图像处理的塑料膜生产缺陷检测方法及系统
CN105469105A (zh) 一种基于视频监控的香烟烟雾检测方法
CN109167928A (zh) 基于显示面板缺陷检测的快速自动曝光方法及系统
CN104101600A (zh) 连铸坯断面微小裂纹检测方法及装置
CN111754466A (zh) 传输机皮带损伤状况的智能检测方法
CN112950541B (zh) 一种基于机器视觉的合成革涂胶在线监测方法
Nandi et al. Machine vision based automatic fruit grading system using fuzzy algorithm
CN113706566B (zh) 一种基于边缘检测的加香喷雾性能检测方法
CN109975301A (zh) 一种用于炼铁高炉的筛网破损检测方法及检测装置
CN108280838A (zh) 一种基于边缘检测的夹片牙型缺陷检测方法
CN115861294B (zh) 一种基于计算机视觉的混凝土生产异常检测方法及装置
CN117928647A (zh) 一种食品加工生产线的品质自动检测方法及系统
CN107657620A (zh) 一种带纹理的金属凝固区域识别的方法及系统
Hashmi et al. Computer-vision based visual inspection and crack detection of railroad tracks
CN108982544A (zh) 一种印制电路板瑕疵部件检测方法
JPH07333197A (ja) 表面疵自動探傷装置
CN118501177A (zh) 一种化成箔的外观缺陷检测方法及系统

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21949043

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21949043

Country of ref document: EP

Kind code of ref document: A1