| Optical non-contact measurement is an important technical means to obtain the three-dimensional shape of the surface of the object.Among them,among the many 3D reconstruction technologies,the active light three-dimensional measurement technology has become the focus of current research.However,in the actual measurement process,because the non-lambertian surface has the characteristics of mirror reflection,global reflection and local multiple reflection,it is easy to form image saturation area and low contrast dark area,which makes the feature information missing and makes it impossible to obtain the three-dimensional information of the area.As a result,three-dimensional measurement of high reflective surface has become a research difficulty in the field of three-dimensional reconstruction of structured light.For this reason,a strip edge based structured light coding strategy is proposed to generate projection patterns in order to solve the problem of high reflectivity of the measured surface.The main contents of this paper are as follows:1.The method of combining Gray code and line shift is adopted as the encoding method of the structured optical system,and the sub-pixel precision stripe edge detection is adopted for the coding pattern to realize high-precision decoding of the encoded image.2.Two sets of structured light systems are built,and the existing calibration methods of structured light are analyzed.The projector in structured light system is regarded as a "pseudo camera",and the calibration of structured light system is realized by Zhang Zhengyou's calibration method.It includes camera calibration and projection calibration.3.In the phase of fringe adjustment,a structured light coding strategy based on stripe edge is proposed to modify the projected fringe image and generate a new projection pattern.In one measurement,this paper proposes a dual exposure fusion strategy,which takes structured light fringe images at high and low exposure times,and fuses them into a group of new coded fringe images with high signal-to-noise ratio.Among them,high exposure is to obtain high quality images of dark areas on the surface of objects,while low exposure is to take high quality images for high reflection areas.In addition,based on the image captured by the camera,the intensity of the saturated pixels corresponds to the mapping area in the projection pattern.Specifically,first of all,in the case of low exposure,the intensity relationship between the projection pattern and the camera captured image is estimated by projecting a series of unifonn gray-scale images onto the high-reflective surface and fitting the non-linear function.Secondly,the threshold intensity mask is set,and the mapping region corresponding to the projection pattern is calculated by using the calibration parameters of the structured light system,and the intensity of the mapping region in the projection pattern is adaptively adjusted based on the estimated intensity relationship.Finally,the two exposure time captured images are fused into a set of fringe images with high signal-to-noise ratio,and the new images are decoded to obtain the corresponding relationship,so as to achieve accurate 3D measurement.4.After the decoded data and calibration data are obtained,the 3-D point cloud data on the surface of the measured object are calculated by using the principle of triangular ranging.The real face and arm are selected as experimental objects to verify the robustness and effectiveness of the proposed algorithm.In addition,in order to verify the accuracy,validity and robustness of the proposed method,three metal objects with high reflectivity were selected for experiments,and the work of several other researchers was restored for comparative experiments.The experimental results show that the structured light technology proposed in this paper has a high measurement accuracy for the object surface with a wide range of reflectivity changes,even for the stainless steel surface. |
PDF Full Text Request