Method for detecting printing carbon line defects of humidity sensor

Abstract

The invention relates to the field of humidity sensor defect detection, in particular to a method for detecting printing carbon line defects of a humidity sensor, which comprises the following steps of: acquiring a clear image of the humidity sensor, and performing threshold segmentation, morphological processing, edge detection, tilt correction, ROI region extraction, histogram accumulation, wavelet transformation and the like on the image to obtain the printing carbon line defects of the humidity sensor. Then, the processed image feature data are extracted to train a BP neural network and a fisher linear classifier respectively to establish a new classifier, and the BP neural network classifier classifies printing carbon line defects and unfilled corners of the humidity sensor. According to the method, the ceramic substrate printing carbon line defects of the humidity sensor can be rapidly and stably detected with high detection precision and classified.

Description

technical field [0001] The invention relates to the field of defect detection of humidity sensors, in particular to a method for detecting defects of printed carbon lines of humidity sensors. Background technique [0002] Humidity sensor is a common tool used to test the moisture content in the air. It is widely used in all walks of life in life. Among them, the polymer resistance type humidity sensor is one of the most widely used humidity sensors. In the production and manufacture of humidity sensors, since the carbon wire circuit is printed on multiple humidity sensor substrates on a whole ceramic substrate, it is necessary to cut the ceramic substrate printed with carbon wires, and then the ceramic sheet After cleaning and soldering the pins on the pad, the next step is to use a polymer solution to dip-coat the humidity sensor ceramic sheet and dry it to form a polymer moisture-sensitive film. [0003] In the production process of the resistive polymer humidity sensor, ...

AI Technical Summary

Problems solved by technology

At present, there are two main aspects of printed circuit detection in the production of polymer humidity sensors. One is performance testing, which needs to check the electrical characteristics of the sensor. This method is time-consuming. The product is visually inspected to determine whether it is defective

However, this kind of manual detection has great limitations, not only the efficiency is low, but also the detection accuracy and yield rate are not high.

Images