Defect detection method and system applied to receiver

Abstract

The invention relates to a defect detection method and system applied to a receiver. The method comprises steps that S1, defect area segmentation, a training sample is inputted through a debugging module, and area segmentation for the training sample is carried out to acquire defect areas; S2, defect characteristic extraction, a defect characteristic description method based on characteristic descriptors is utilized to carry out characteristic extraction of characteristics of the defect areas according to the defect areas acquired in the step S1; S3, defect target identification, the defect characteristics are utilized, BP neural network training is carried out through sample and parameter selection to acquire classification rules; and S4, online detection, pictures are acquired from a production line at each time interval, the classification rules are utilized to discriminate the acquired pictures in real time to acquire a detection result, and the data information including the yield data in each production process is further collected. The method is advantaged in that not only can detection efficiency is improved, but also production conditions are mastered integrally, production can be improved through utilizing the acquired detection information, and enterprise revenue is improved.

Description

technical field [0001] The invention belongs to the text input technology of electronic equipment, in particular to a defect detection method and system applied to a receiver. Background technique [0002] The receiver is mainly composed of three parts: the vibration part, the magnetic circuit part, and the cavity part. Its production includes the manufacture of the vibration part, the manufacture of the magnetic circuit part and the assembly of the cavity. Each part includes multiple manufacturing processes. The structure of the receiver itself, the production process and the final testing process will have an impact on the performance and pass rate of the receiver. Therefore, in order to achieve a high pass rate, the structure of the receiver must be optimized. Production process and Improve detection efficiency and accuracy. [0003] In order to achieve the above technical problems, people have carried out long-term exploration. For example, a Chinese patent discloses a ...

AI Technical Summary

Problems solved by technology

[0005] The above two solutions solve some problems in the prior art from different angles, but none of them involves the detection process of the receiver in the production process, and the detection process is also very important as the last process in the production process. Among them, the quality detection of the vibration part is the most difficult

In the prior art, the quality inspection of the vibration part of the receiver mainly adopts the method of manual inspection. By placing the vibration plate voice coil face up in the turnover box, the defective products are eliminated. According to the defective phenomenon of the vibration plate, it is fed back to In the previous manufacturing process, process improvement, manual inspection is not only inefficient, but also has low detection accuracy, so it is difficult to meet the long-term development of the enterprise

Moreover, it takes about 20 seconds to manually detect the defects of the receiver winding and voice coil coating, and the correct detection rate is only about 96%, and it is often impossible to accurately measure the defect area and find out the production problem.

Images