Simulation of a machine vision system for reflective surface defect inspection based on ray tracing

Abstract

A complete simulation of a machine vision system aimed at defect inspection on a reflective surface is proposed by ray tracing. The simulated scene is composed of the camera model, surface reflectance property, and light intensity distribution along with their corresponding object geometries. A virtual reflective plane geometry with scratches of various directions and pits of various sizes is built as the sample. Its realistic image is obtained by Monte Carlo ray tracing. Compared to the pinhole camera model, the camera model with a finite aperture emits more rays to deliver physical imaging. The bidirectional reflectance distribution function is applied to describe the surface reflectance property. The illustrated machine vision system captures a number of images while translating the light tubes. Then the image sequence obtained by experiment or simulation is fused to generate a well-contrasted synthetic image for defect detection. A flexible fusion method based on differential images is introduced to enhance the defect contrast on a uniform flawless background. To improve detection efficiency, defect contrast of synthetic images obtained by various fusion methods is evaluated. Influence of total image number, light tube width, and fusion interval is further discussed to optimize the inspection process. Experiments on car painted surfaces have shown that the simulated parameters can instruct the setup of the optical system and detect surface defects efficiently. The proposed simulation is capable of saving great effort in carrying out experimental trials and making improvements on reflective surface defect inspection.

Optical Material Characterisation of Prepreg CFRP for Improved Composite Inspection

“…Furthermore, they analysed several simulation techniques for calculating the diffraction parameters based on the scattering angle. Additionally, Zhang et al [46] presented a method to simulate an image processing system for the inspection of a reflecting surface by means of ray tracing. The overall simulation consists of a defined camera model in combination with the reflection properties of the surface as well as the intensity and distribution of the reflected radiation as input parameters.…”

Robot-Inclusive False Ceiling Design Guidelines

“…This is especially so if the robot utilises point light sources for illumination or if the light source is located near to the robot's camera itself. Moreover, defects on reflective surfaces become only detectable from a particular direction from the angle of illumination [64]. Using matte surfaces for the amenities, as well as an ambient or diffuse light source, would help improve surface defect detection.…”

Light-emitting diode planar array light source system for defect detection of highly reflective surface

“…Yong-Ju Jeon et al [1] proposed the method of using dual-light switching lighting, using alternating black and white patterns to detect defects on steel surfaces. Zhang et al [2] proposed a machine vision system based on ray tracing, which translates the lamp array to shoot a large number of images and generates high-contrast detection images through image fusion. Feng et al [3] proposed a fast calibration method of camera response function based on high dynamic range imaging technology, adjusted chroma information through tone mapping algorithm, reduced local overexposure from the algorithm, and improved the clarity and contrast of synthetic images, but this method could not eliminate the influence of overexposure area.…”