Due to the problem of insufficient dynamic range， which may cause image supersaturation and low fringe contrast， traditional fringe projection three-dimensional （3D） measurement technology makes it difficult to obtain complete and effective 3D profiles of objects with non-Lambertian surfaces. To enhance the adaptability of structured light 3D measurement， we proposed an improved high dynamic range （HDR） 3D measurement method using line-shifting strips， which were more robust than sinusoidal fringes， as encoding patterns. A strip contrast model of complementary line-shifting strip sequences was constructed for optimum contrast detection of different color channels responding to light intensity. Subsequently， a color camera could be used to capture a color image sequence in a single exposure time to synthesize new line-shifting strip images， avoiding invalid or error data introduced by insufficient dynamic range. Finally， by using background normalization to unify the background of different channel components of the synthesized image， the line positions could be precisely located through complementary line-shifting strips. This allowed for the accurate demodulation of line encoding information， enabling 3D reconstruction. Experimental results demonstrate that， compared to the multiple exposure method， the image acquisition time is reduced by 82% and the total reconstruction time is reduced by 59%. The proposed method can demodulate the effective line encoding information of metal workpieces with complex surface reflection characteristics to obtain the complete and high-accuracy 3D point cloud， achieving efficient high dynamic range 3D measurement.