Massive concrete structures are widely used in construction projects, but their quality inspection faces technical bottlenecks. This study proposes a single-side imaging technique for mass concrete structures based on 3D laser scanning technology, which achieves accurate characterization of concrete structures through 3D laser point cloud acquisition, wavelet transform denoising and implicit surface reconstruction. The study firstly adopts 3D laser scanning technology for image acquisition, and applies the seed point 4-neighborhood grid distribution method to obtain the original image estimation; secondly, empirical mode decomposition is applied to measure the information of the scanned image; then wavelet transform processing is utilized to remove the noise and improve the quality of the data; finally, hole repair and surface reconstruction are realized through resampling, eigenface extraction and implicit surface construction. The experimental results show that, compared with RIEGL VUX-1UAV, the accuracy of this method is comparable in elevation difference measurement (the mean value of elevation difference is 0.019), but the point cloud density is improved by 2.75 times on average, and the average thickness of the point cloud at the edges is only 5.32 cm. Meanwhile, the method distributes the point cloud uniformly in the different thickness intervals, and the average proportion of the point cloud reaches 0.216, which ensures that the strength of the concrete structure is later stable growth, so that the strength assurance rate reaches 98.9% at a probability degree of 2.3. This technique provides an efficient and accurate solution for the inspection of mass concrete structures and can be widely used in the field of engineering quality assessment and monitoring.
