基于数字图像技术的碎石撒布情况检测方法研究

Study on Detection Method of Gravel Spreading Based on Digital Image Processing Technology

近年来,随着数字图像处理技术的兴起,越来越多的路段在道路建设过程中使用数字图像处理技术对道路建设情况进行检测。为了在同步碎石封层施工阶段对碎石撒布率和撒布均匀性进行实时评价,实时反馈碎石撒布情况并指导下一步的施工作业,减少不满足规范要求的撒布路段,在对撒布路面碎石图像经过图像灰度化、小波高频系数改进阈值处理、低频系数改进Retinex法处理、小波重构及二值化处理等数字图像处理后,得到图像碎石撒布率,并统计了图像中横向、纵向及四分域各部分碎石撒布面积的分布得到均匀性系数。通过对施工现场采集图像的撒布率与传统试验方法得到的结果进行比较,验证了所提算法的准确性并提出了数字图像碎石撒布均匀性评价指标。试验结果表明:通过数字图像处理技术得到的碎石撒布率与实际撒布率检测差值在±1%内,能够满足规范检测精度要求;通过采用四分、横向及纵向分析方法,得到的各项均匀性系数可用于评价路面撒布均匀性并确定撒布情况较差的具体位置,能够对下一步的碎石撒布施工作业进行指导,为同步碎石封层施工提供了一种快速、无损、有效和准确的数字图像检测方法。

In recent years, with the rise of digital image processing technology, more and more road sections use digital image processing technology to detect road construction in the process of road construction. In order to evaluate the spreading rate and uniformity of gravel in real time during the construction stage of the synchronous gravel seal, feedback the gravel spreading in real time and guide the next construction operation, and reduce the spreading sections that do not meet the requirements of the specification, digital image processing such as image grayscale, wavelet high-frequency coefficient improvement threshold processing, low-frequency coefficient improvement Retinex method processing and wavelet reconstruction and binarization processing are performed on the images of spread gravel on pavement, the image gravel spreading rate is obtained, and the distribution of spread gravel in the horizontal, vertical and quartile domains in the image is counted to obtain the uniformity coefficient. The accuracy of the proposed algorithm is verified by comparing the spreading rate of the images collected at the construction site with the result obtained by the traditional test method, and an evaluation indicator for the uniformity of the gravel spreading in digital image is proposed. The test result shows that(1) the detection difference between the distribution rate of gravel obtained by digital image processing technology and the actual distribution rate is within ±1%, which can meet the requirements of standard detection accuracy;(2) by using quartile domains, horizontal and vertical analysis method, the obtained uniformity coefficients can be used to evaluate the uniformity of pavement spreading and determine the specific location with poor spreading conditions, which can guide the next construction of gravel spreading and provide a fast, non-destructive, effective and accurate digital image detection method for the synchronous gravel seal construction.