Continuous casting slab surface crack depth measurement using sinusoidal phase grating method

Abstract

With the development of hot charging and direct rolling technology in continuous slab casting, there are increasing demands for the surface quality of hot continuously cast slab evoked by modern metal material fabrication. In order to reduce noise, which is caused by scale, water film and oscillation marks, during on-line detection of surface defects, a sinusoidal phase grating projection method is presented to detect the depth and surface profile of surface cracks in continuous slab casting. This technology, based on a deformed projected grating pattern, which carries three-dimensional information of the measured surface of the slab, can provide an absolute measurement of the depth of the surface defect by using a temporal phase unwrapping algorithm. The main advantage is that objects containing depth discontinuities of the crack are profiled as easily as smooth ones. The experimental results show the method has a high signal to noise ratio, and good applicability for automatic crack depth testing in continuous casting.

Research Progress of Automated Visual Surface Defect Detection for Industrial Metal Planar Materials

“…Grating projection technology is a kind of widely used structured light that actually belongs to the generalized surface structured light. Grating projection technology is actually a kind of surface structured light in a broad sense, which generates sinusoidal fringe through computer programming, projects the sinusoidal fringe to the measured object through the projection device, uses a CCD camera to shoot the bending degree of the fringe adjusted by the object, demodulates the bending fringe to get the phase, and then converts the phase to the height of the whole field [ 158 , 159 ]. Although the active structured light approach contains many advantages (for example, high spatial resolution and accuracy), it embodies several problems: Firstly, active structured light usually requires spatial or temporal phase unwrapping, which increases the computational complexity and reduces the detection speed.…”

Review of surface defect detection of steel products based on machine vision

“…Zhang et al. [24] used the sinusoidal phase grating projection method by integrating a grating projector and a CCD camera to measure the depth of continuous casting slab surface cracks. Wen et al.…”

“…5 m/s. Zhang et al [24] used the sinusoidal phase grating projection method by integrating a grating projector and a CCD camera to measure the depth of continuous casting slab surface cracks. Wen et al [25] installed two CCD cameras on both sides of the digital light processing (DLP) projector with blue light coded patterns, and used stereo rectification and column-coded patterns to obtain the surface height information of high-temperature steel products.…”

Surface defect detection of steel strips based on classification priority YOLOv3-dense network

“…Luo et al employed the selectively dominant local binary patterns (SDLBPs) algorithm to classify the surface defects of hot-rolled strips so as to obtain higher classification accuracy and time efficiency, yet they failed to achieve target defect detection [7]. Zhang and other scholars proposed a sinusoidal phase grating projection method to detect the depth and surface profile of cracks in continuous slab casting, which is suitable for the detection of defects on the surface of the slab [8]. Additionally, the YOLO network has also been used to detect the surface defects of steel strips.…”