Industrial image uniform-illumination processing method

Abstract

The invention provides an industrial image uniform-illumination processing method comprising: obtaining an industrial image in nonuniform illumination by a camera and subjecting the industrial image to gray processing; and subjecting the gray image to wavelet decomposition to obtain the first high-frequency component and the first low-frequency component of the industrial image; removing the first high-frequency component of the industrial image and subjecting the first low-frequency component of the industrial image to second-time wavelet decomposition, and in the same way, using the eighth low-frequency component of the industrial image as the illumination estimated map of the industrial image; making subtraction of the illumination estimated map and the gray image to obtain a background difference map of the industrial image; and adding the mean values of the background difference map and the gray image to obtain background light-uniform map of the industrial image. The method has good adaptability, and has a good treatment effect for nonuniform illumination in different situations.

Description

technical field [0001] The technical field of industrial image detection of the present invention, in particular, relates to a method for uniform light processing of industrial images. Background technique [0002] Because the images on the industrial assembly line move quickly on the assembly line, the collected images are often distorted and positionally shifted, and the collected images are not easy to detect. On the basis of perfect hardware conditions, the position deviation of the image can be controlled. The shift does not exceed ±2mm, but image distortion and a certain degree of positional offset are impossible to avoid, which need to be considered in industrial image detection. [0003] In the prior art, in the industrial image detection, due to environmental reasons, it is necessary to increase the artificial light source to obtain a clearer sample image, but this also brings disadvantages at the same time, that is, the artificial light source illuminates the sampl...

AI Technical Summary

Problems solved by technology

[0002] Because the images on the industrial assembly line move fast on the assembly line, the collected images tend to be distorted and position shifted, and the collected images are not easy to detect. On the basis of perfect hardware conditions, the position deviation of the image can be controlled. The shift does not exceed ±2mm, but image distortion and a certain degree of positional offset are impossible to avoid, which need to be considered in industrial image detection

[0003] In the existing technology, in industrial image detection, due to environmental reasons, it is necessary to increase the artificial light source to obtain a clearer sample image, but this also brings disadvantages at the same time, that is, the artificial light source illuminates the sample measurement, resulting in uneven background The image, that is, the image with a bright center and a dark surrounding, the current uniform light method is mostly to improve the hardware conditions

It will increase the cost, and at the same time, it has no adaptability, and can only adapt to a single environment. If the environmental conditions change slightly, the hardware conditions also need to be changed.

Images