Laser Image Processing Non-Destructive Testing Method for Soil Compaction

In order to achieve soil compaction detection, a laser image measurement system for soil compaction was established. In order to study the subgrade compaction degree by laser image, the roadbed soil samples with different compactness were laser-imaged in the laboratory through computer vision system, and the image gray scale was obtained through image processing, and the correlation was established with the compaction degree. The equations were then tested in the field to give the compaction of the subgrade. The test results show that it is feasible to detect the compaction of the roadbed by laser image, and it will not damage the roadbed main body. Firstly, the soil laser image is collected and smoothed by 4 neighborhood averaging method. Secondly, the canny algorithm is used to extract the laser spot in the laser image. Then the water content, laser spot radius, absorption coefficient and scattering coefficient are selected as the classifier. The input characteristic parameters; finally, the back propagation (BP) neural network is used to predict the degree of compaction. The experimental results show that the BP neural network meets the measurement accuracy requirements after 11 learning; the average absolute and relative errors are about 2% compared with the actual measured values of the ring cutter method. Therefore, the detection accuracy of the measurement system of this paper meets the detection requirements of soil compaction.