Abstract: To enhance the anti-interference capability and robustness of visual measurement technology in complex road environments and achieve precise measurement of three-dimensional texture information on road surfaces, the dual-spectrum sinusoidal encoded fringe structured light was introduced, and a second-order fusion binocular vision measurement model under dual constraints of dual-spectrum and structured light was constructed. Complex lighting conditions and road environments were simulated to evaluate the measurement effectiveness of the improved algorithm in complex environments. The results indicate that, through T-test analysis, there is no significant difference in measurement accuracy between the binocular fusion structured light vision algorithm and the skid resistance texture tester at a 0.05 confidence level. The introduction of dual-spectrum structured light constraints effectively improves the resilience of binocular vision technology against complex environments. Meanwhile, the binocular fusion structured light vision algorithm can maintain good measurement stability in complex lighting conditions ranging from 100~1 250 Lux, as well as in complex road environments such as mottled surfaces, water film, and oil stains. Guided by the principles of informatization and intelligence, the research results are of great significance for enriching the research scope of road surface functions and creating a comfortable and safe traffic environment.

Key words: highway engineering; binocular vision; structured light; resistance to complex environment

摘要： 为了提高视觉测量技术在复杂路用环境下抗干扰能力和鲁棒性，实现复杂环境下路面三维纹理信息的精确测量，引入双光谱正弦编码条纹结构光，构建双光谱和结构光双重约束下二阶融合双目视觉测量模型，构造复杂光照和复杂路用环境，开展复杂环境下改进算法的测量效果评价。结果表明：经T-test分析，双目融合结构光视觉算法的测量精度与抗滑纹理测试仪在0.05的置信水平下不存在显著差异，双光谱结构光约束的引入能够有效改善双目视觉技术抗复杂环境的能力。同时，无论是在100~1 250 Lux复杂光照变化下，还是在斑驳、水膜和油污等复杂路用环境下，双目融合结构光视觉算法均能保持良好的测量稳定性。

关键词: 道路工程；双目视觉；结构光；抗复杂环境