A method for detecting surface defects of magnetic circuit materials

Abstract

The invention discloses a method for detecting surface defects of magnetic circuit materials, which comprises the following steps: collecting training pictures; manually labeling; training the training pictures and marked pictures with a designed convolutional neural network; inputting the data collected by a camera The prediction result is obtained from the unlabeled image of the magnetic circuit; the detection result is generated according to the preset threshold fusion. The method provided by the invention can self-learn and optimize the network model of the defect form on the surface of the magnetic circuit product, and overcome many adverse effects caused by interference factors such as random product defect form, complex texture, and low contrast, especially when a small number of sample inputs Under the condition of using data enhancement, an excellent network model can also be obtained, thereby improving the defect recognition rate. In addition, the data processing speed of the method of the present invention is at least equivalent to that of the existing machine vision algorithm, and the accuracy and yield rate far exceed the existing machine vision. Visual algorithms can significantly improve the factory quality of magnetic circuit products.

Description

technical field [0001] The invention relates to a detection method of a magnetic circuit material, in particular to a method for detecting surface defects of a magnetic circuit material based on a deep learning algorithm. Background technique [0002] The magnetic circuit is mainly used as the component material of the mobile phone speaker in the 3C industry. The quality of the magnetic circuit in the production process directly affects the acoustic performance of the speaker. Usually, due to defects such as cracks, dirt, impurities, and lack of appearance on the surface of the magnetic circuit, manual online judgments are used in traditional production lines, but manual work also causes product yields to decline due to fatigue and human eye limitations, which affects The quality of shipments cannot meet the requirements of end users. [0003] In recent years, some system integrators have adopted machine vision methods to detect magnetic circuit surface defects. First, ima...

AI Technical Summary

Problems solved by technology

Due to the complex surface texture of the magnetic circuit, the surface defect shape is very random, and the contrast is low. At the same time, the stability of the magnetic circuit imaging on the production line is also difficult to guarantee. The existing machine vision algorithm has great limitations in the process of analyzing defects. High rate and misjudgment rate, it is difficult to meet the actual production requirements

Moreover, the existing machine vision algorithms need to constantly adjust and optimize the algorithm when facing various random defects, which has very poor adaptability and will also lead to a waste of time.

Images