Fine Structural Characterization of Construction Asphalt Mixtures Based on Image Processing Techniques

The development of economy is inseparable from the construction of traffic buildings. Especially in the current road construction, asphalt mixture is mainly used for pouring. Over time, the asphalt mixture is disturbed by other external factors, resulting in a decrease in the performance of the asphalt mixture. Under this background, this paper mainly studied the mesoscopic angle of the construction asphalt mixture through image processing technology, and analyzed the mesoscopic structural characteristics of the construction asphalt mixture. This paper took the void structure as the research index, and performs image enhancement, image denoising, image sharpening, image segmentation and image edge detection on the collected images of building asphalt mixture in turn. In terms of image enhancement, the image after histogram equalization is clearer in texture, distinct in layers and more prominent than the original image. In terms of image denoising, the median filter method is used, and the noise reduction effect is obviously better than other methods. In terms of image sharpening, the contour of the image sharpened by the Laplacian operator is clearer. In the aspect of image segmentation, the threshold segmentation method has obvious image void boundary and detail information, which is conducive to extracting void information. In terms of edge detection, the image lines under the Canny operator are complete, which greatly reduces the loss of edge information. On this basis, the void structure model was constructed and tested experimentally. The results showed that the average equivalent diameter, average perimeter, and average contour area of the voids in each layer have roughly the same trends as the layers increase. Not only that, the detected void ratio was about 8.14%, which was only 1.17% different from the actual void ratio. This showed that the void structure model constructed under the image processing technology has a significant effect on the porosity detection, and this result brought certain guiding suggestions for the follow-up study of mesostructure characteristics.