"Machine Vision Inspection" What are the advantages of machine vision in text quality inspection?

During the printing process, experienced printing workers use density meters and other measuring tools to test the finished products, which can improve the printing quality of the product and reduce the scrap rate. But in high-speed printing production, such detection means a waste of time. At the same time, when giving evaluations, different personnel have different control of quality standards, which will also lead to inconsistent, unstable, and inaccurate testing and evaluation results.

Under this situation, in order to achieve rapid detection of printing quality and improvement of objective evaluation indicators, in view of the limitations of human eyes in detecting and evaluating text printing quality, the industry has introduced machine vision technology, image processing pattern recognition and other means to detect and evaluate text printing quality, which has greatly solved this problem.

The text printing quality detection and evaluation system generally consists of five parts: text image capture card, captured image quality assessment, image preprocessing, feature extraction and quality evaluation. In order to ensure that the collected images have high resolution, the text printing detection technology uses a 5-megapixel industrial digital camera to capture images of printed matter, then extracts its information features and converts them into identifiable codes, and further combines these codes into feature templates. During quality evaluation, the recognition system obtains its characteristic template and compares it with templates that already exist in the database. If the standard is met, printing will continue. If the standard is not met, the machine will stop for inspection. The identification and detection process is shown in the figure.

Based on the machine vision system, the coded text can also be recognized and detected. First, it extracts features such as font outline, grid, and projection, and then makes a judgment based on statistical features. Finally, if the font outline features are particularly close, local features of the inner edge are extracted for a second judgment. This method has achieved good results in industrial applications. There are many other applications based on machine vision technology in text printing, character detection, and inkjet text defect detection, which fully reflect the superiority of machine vision technology.