Polished tile defect detection method based on multi-layer feature extraction and related equipment

Abstract

The invention discloses a polished tile defect detection method based on multi-layer feature extraction and related equipment, and the method comprises the steps: collecting a first image of a to-be-detected polished tile under a preset environment condition; inputting the first image into a trained recognition network model; obtaining the output of the identification network model as the defect type of the to-be-detected polished tile; wherein the identification network model comprises a feature extraction sub-network used for extracting deep features of an image input into the recognition network model, and a recognition sub-network used for recognizing and judging the deep features extracted by the feature extraction sub-network. The feature extraction sub-network adopts a feature pyramid form and a multi-layer feature extraction method to perform deep feature extraction on an input image, and the recognition sub-network works by adopting an attention mechanism. According to the invention, automatic identification of the defect type of each polished tile can be realized, the defect type identification precision is improved, the automation degree is high, the automatic detection efficiency is high, and the invention can be widely applied to the polished tile production industry.

Description

technical field [0001] The invention relates to the field of computer processing, in particular to a method, system and equipment for detecting defects of polished tiles based on multi-layer feature extraction. Background technique [0002] With the development of society, various industrial products are used more and more widely. As an important building and decoration material, polished tiles have been widely used and have a huge market. In the production process of polished tiles, due to the influence of various factors such as process and environmental parameters, there may be various types of flaws in polished tiles, and the types of flaws that may be caused by different factors will be different. In the production process, in order to ensure the yield of polished tiles, it is necessary to detect the type of defects in time, so as to adjust the relevant parameters of the polished tiles production process. At this stage, the flaw detection of polished tiles in the marke...

AI Technical Summary

Problems solved by technology

At this stage, the flaw detection of polished tiles in the market relies on manual operations, which has high work intensity, low manual detection efficiency, and poor detection accuracy.

Images