Research On Three-dimensional Topography Measurement Method Of Highly Reflective Objects Based On Structured Light

3D measurement technology is more and more widely used in the fields of cultural relic restoration,reverse engineering,culture,education and art,Color structured light technology based on stripe coding is the mainstream method of 3D measurement due to its advantages of high efficiency and low cost.However,when measuring objects with highly reflective surfaces such as ceramics and metal parts,there will be high light spots in the highly reflective areas on the surface of the object,resulting in the loss of stripe information.At the same time,the stripes near the highly reflective areas will appear color distortion,affecting the center of the stripes.Extraction precision.In view of the above problems,this paper improves the existing specular surface measurement methods,and proposes a specular surface structured light measurement method based on adaptive projection intensity.The main research contents are as follows:(1)In the process of using the structured light method to measure objects with highly reflective surfaces,the projected fringe pattern will be broken due to the high light points on thf surface of the object,resulting in the moss of fringe information collected by the camera and affecting the accuracy of the measurement.Aiming at the above problems,this paper proposes an adaptive projection intensity calculation method.First,judge the reflective characteristics of the surface of the object according to whether there is a highlight area on the surface of the object after projection.For the object judged to be a highly reflective surface,first select the maximum,minimum and half of the maximum projection intensity for projection,and then calculate the first Calculate the adaptive projection intensity of the round projection,and count the area of the highlight area in the results collected in the first round,fit the relationship between it and the projection intensity of the first round,and calculate the adaptive projection intensity of the second round of projection according to the fitting results.And re-projection,so that each time the projection intensity is modified,the variation of the highlight area area approaches a fixed value.The experimental results show that the method can accurately obtain the optimal projection intensity in the highlight area,and effectively remove the fringe missing phenomenon in the highlight area.(2)In the process of using color structured Light to measure highly reflective surface objects,due to the color distortion of the stripes,the extraction accuracy of the fringe center is not good and a set of color structured light encoding and decoding methods is proposed.During encoding,the uniqueness of the stripe combination is guaranteed by constructing a DeBruijn sequence,and the stripe color is constrained and selected in the HSV color space;when decoding,the V channel of the image is first repaired using the multi-scale Retinex method to eliminate the brightness of the stripe image.Then,perform first-order differential filtering on the restored V channel image,obtain the coordinates of the center point of the stripe according to the obtained waveform,and then restore the color of the center paint of the stripe through the coordinate index and threshold constraint to obtain a complete color.Stripe center result.The experimental results show that the codec scheme can accurately obtain the center of the fringe image on the highly reflective surface.In older to verify the validity and accuracy of the method proposed in this paper,four highly reflective materials such as ceramics,silver,metallic paint and plastic leather were measured using the method proposed in this paper.The measured models included spheres.cylinders,teapots,etc.The experimental results show that the missing parts of the highlight area on the surface of the object are well restored in the model,and the reconstructed model results are relatively smooth.After calculation,the maximum relative error of the measurement results is 3.875%,and the minimum relative error is 0.385%.The error is 2.214%,which confirms that the method proposed in this paper has high accuracy,and has strong adaptability to different materials and models,and can meet the needs of daily consumer-level measurement.