Adaptive Coding Fringe Projection Profilometry on Color Reflective Surfaces
Wang Ying,Ni Yubo,Meng Zhaozong,Gao Nan,Guo Tong,Yang Zeqing,Zhang Guofeng,Yin Wei,Zhao Hongwei,Zhang Zonghua
DOI: https://doi.org/10.3788/aos231894
2024-01-01
Acta Optica Sinica
Abstract:Objective Fringe projection profilometry is widely employed to reconstruct the three-dimensional(3D)shape of an object surface.However,when this method is utilized to measure objects with color reflective surfaces,the image captured by the camera is oversaturated with pixels due to ambient lighting and reflections from the projected fringes,which results in the inability to measure the surface of the reflective area.This problem is mainly due to the unevenly varying reflective of surfaces,which is affected by both the roughness and the surface color.To solve the problem of eliminating the interference of the object surface color and complete the 3D shape measurement method based on the reflectivity change of colored highly reflective surfaces,we propose an adaptive generation of complementary color sinusoidal fringes method.By different absorption of colors by the object surface color to be measured,a complementary color of lighting is projected onto the highly reflective area to reduce the surface reflectivity of the region and suppress the exposure phenomenon. Methods We put forward a method to measure the 3D shape of colored objects with high reflectivity,which is based on adaptively encoded complementary color fringes.Firstly,the highly reflective region of the object to be measured should be located.The image of the object surface is captured by the camera when the projector projects the strongest white light,and the coordinates of the oversaturated pixel points are extracted by an inverse projection technique.The location of the highly reflective region in the coordinate system of the projected image is obtained via the matching relationship between the projector and the camera.Then,the optimal color adopted for projecting the highly reflective region of the object is calculated by the color image of the object surface and then captured by the camera.The projecting color obtained in the previous step is employed to generate an image that is projected to the highly reflective region on the measured surface.The saturation value of the adopted projecting color is adjusted according to the magnitude of the adjacent light intensity values at either end of the boundary encoded color until the adjacent light intensity values are less than 20.Finally,after sinusoidal fringes on the V component of the HSV color space are encoded,and meanwhile adaptive complementary color sinusoidal fringe patterns are generated and projected onto the object surface to be measured.The complete 3D shape of the object surface to be measured is recovered by solving the unwrapped phase. Results and Discussions The proposed method employs adaptively encoded complementary color fringes.It reduces the reflectivity of the highly reflective region on the surface,solves the unwrapped phase loss after utilizing traditional fringe projection profilometry,and finally obtains the complete 3D shape of the yellow ceramic cup(Fig.5).Additionally,the phase resolution results of the yellow ceramic cup are compared and analyzed by traditional gray fringes and the proposed complementary color-coded fringes under different exposure time.The results show that when the exposure time is greater than 40 ms,the phase recovery completeness of the region D is maintained at 100％(Fig.6)by applying the proposed method.The purpose of measuring the complete 3D shape of the surface of a color highly reflective object by projecting only a set of adaptively encoded complementary color sinusoidal fringe patterns is achieved.Meanwhile,the mean error of the proposed method is 0.5281 mm,smaller than that of the traditional multiple exposure method.In conclusion,this method is not only more efficient than the traditional multiple exposure method in the measurement process but also improves the measurement accuracy. Conclusions To address the challenges in measuring the 3D shape of colored highly reflective objects,we propose a novel fringe projection profilometry method based on adaptive color encoding.The proposed method encodes and projects fringe structured light complementary to the measured surface color into the highly reflective region in the HSV color space based on the theory of photometric complementarity.As a result,it reduces the surface reflectivity of the highly reflective region and achieves 3D shape measurement of colored highly reflective objects.The experimental results show that this method reduces the number of projected images during the measurement compared with the traditional multiple exposure methods.Only a set of adaptively encoded complementary color sinusoidal fringe maps should be projected to obtain a complete 3D shape of the surface of a colored highly reflective object.The proposed method shows certain advantages in measurement efficiency and accuracy.