Tire X-ray image impurity defect detection method

Abstract

The invention relates to a tire X-ray image impurity defect detection method. The method comprises the following steps: S1, establishing an image data set containing impurity defects; s2, enhancing the data set by using rotation, mirroring and brightness transformation; s3,building an improved YOLOv4-tiny network; s4, setting hyper-parameters and network parameters for training; s5, inputting a to-be-detected image to carry out defect detection. According to the invention, end-to-end tire impurity defect positioning is realized, the requirement of real-time tire defect detection can be met, and the production efficiency is greatly improved.

Description

technical field [0001] The invention relates to a method for detecting impurity defects in tire X-ray images, in particular to a method for detecting impurity defects in tire X-ray images based on improved YOLOv4-tiny. Background technique [0002] In recent years, with the improvement of living standards, the number of cars in my country has increased year by year, and it is expected to surpass the United States in 2021 to become the country with the largest number of cars in the world. As a necessary means of transportation for people to travel, cars play an indispensable role in people's lives. [0003] As the main part of the car, the tire is also the only part of the car that is in contact with the ground. It is responsible for alleviating impact and carrying the weight of the car. The quality of tires is related to the safety of drivers. At present, domestic tire manufacturers mainly rely on imported foreign X-ray machines to assist manual tire quality assessment. Du...

AI Technical Summary

Problems solved by technology

At present, domestic tire manufacturers mainly rely on imported foreign X-ray machines to assist manual tire quality assessment. Due to the blockade of foreign technology, the corresponding domestic tire defect automatic detection algorithm is not perfect, and a few enterprises still mainly use manual naked eye detection. The efficiency of artificial naked eye detection is low, and it is greatly affected by human subjectivity

Due to the complex background of tire multi-texture anisotropy, the defects have different shapes, and the defects account for less than 1% of the texture background, which poses challenges for the detection of defects.

[0004] The existing tire defect algorithms are mainly traditional defect detection algorithms, which are mainly divided into statistics-based algorithms, frequency-domain-based algorithms and model-based algorithms. At the same time, it is difficult for the algorithm to detect defects and defects with low background contrast, such as the detection of impurity defects located on the tread, and the detection efficiency of the entire tire X-ray image is low, which cannot meet the industrial real-time requirements. testing requirements

Images