In the three-dimensional (3D) scanning measurement based on structured light techniques, the strong reflection surface is easy to produce local specular reflection due to the illumination of the structured light, which will cause the camera to be over-exposed, and therefore the geometry information of strong reflection surface cannot be detected. Since the digital micromirror device (DMD) has the modulating characteristics of the spatial information of incident light, an adaptive high-dynamic-range imaging method based on DMD is proposed to solve the problem of visual imaging of strong reflection surface. Firstly, a novel and computational imaging system is designed and built, and its optical model is also established. Then, the matching and mapping methods between DMD micromirrors and CMOS pixels are described in detail and realized. Meanwhile, we analyze the theory of the high-dynamic-range imaging based on per-pixel coded exposure, and design a coding control algorithm of light intensity to achieve the adaptive precision modulation of the intensity of incident light, so that the incident light in the imaging system is always in appropriate exposure intensity. The experiments show that the method can break through the limited dynamic range of the ordinary digital camera, and accurately control the intensity of incident light in each region of the measured strong reflection surfaces, and thus it can obtain the high-quality images of the local over-exposure area of the strong reflection surface. More importantly, the research will provide a new solution to the problem of 3D point cloud loss caused by local over-exposure of the strong reflection surface.