Solid wood panel quality detection method with combination of computer vision and self-learning behaviors

Abstract

The invention discloses a solid wood panel quality detection method with the combination of computer vision and self-learning behaviors. A quality detection model is constructed, online detection of solid wood panels is achieved, the constructed quality detection model is used for performing quality detection on solid wood panels on a production line, and if the correct recognition rate is not qualified, self-learning correction is performed on the quality detection model so that the correct recognition rate can be qualified. According to the solid wood panel quality detection method, the influence caused by solid wood panel batch differences on quality detection can be avoided, efficient application of the computer vision method in the field of solid wood panel quality detection can be achieved, and the method can be applied to solid wood quality detection.

Description

technical field [0001] The invention relates to a computer vision detection method, in particular to a solid wood plate quality detection method combining computer vision with self-learning behavior, and belongs to the field of wood processing and manufacturing. Background technique [0002] Solid wood boards are durable and have natural textures. Most of them have the unique aroma of natural wood. They have good hygroscopicity and air permeability. They are beneficial to human health and do not cause environmental pollution. They are high-quality boards for making high-end furniture and decorating houses. In the production and processing of solid wood panels, solid wood panels can be divided into defective products and different grades of genuine products according to their own quality characteristics. Defective products generally refer to products that have obvious scars, insect holes, cracks, deformation and other defects on the surface of the board, which seriously affec...

AI Technical Summary

Problems solved by technology

Since the surface characteristics (color, texture, etc.) of solid wood panels are affected by many factors such as wood growth environment, material location, processing technology, tree species, etc., solid wood panels are highly complex non-standard products, and the computer developed for specific batches of solid wood panels Visual inspection programs often cannot be effectively applied to other batches of products, which greatly limits the application of computer vision methods in the field of quality inspection of solid wood panels

Images