Tobacco shred impurity detection method based on Yolo V3 model

Abstract

The invention provides a tobacco shred impurity detection method based on a Yolo V3 model, which comprises the following steps: establishing the Yolo V3 model, and shooting tobacco shreds by using an industrial camera to obtain a tobacco shred image; and performing data processing on the tobacco shred image, collecting and processing an impurity image generated in the tobacco shred production process, and performing enhancement processing by using the two images to form a tobacco shred impurity training data set. And performing K-Means value clustering calculation on the tobacco shred impurity training data set to obtain a new anchor box value, and replacing a data set parameter in the original Yolo v3 model with the new anchor box value. And training a Yolo V3 model through the tobacco shred impurity training data set to obtain an improved Yolo V3 model, and performing impurity target detection on the tobacco shred image by using the improved YolV3 model. The accuracy and the working efficiency of tobacco shred impurity detection can be improved, and the production quality of cigarettes is improved.

Description

technical field [0001] The invention relates to the technical field of cigarette production management, in particular to a method for detecting sundries in shredded tobacco based on the Yolo V3 model. Background technique [0002] During the production process of cigarettes, the shredded tobacco is often inevitably mixed with paper or plastic packaging debris during transportation, which affects the quality of cigarettes. Therefore, it is necessary to detect impurities on the shredded tobacco on the production line. Manual selection depends entirely on human eyes to find impurities and foreign objects in shredded tobacco. It is affected by individual subjective consciousness, and it is time-consuming and laborious. Working for a long time will easily cause visual fatigue and work efficiency cannot be guaranteed. When the traditional machine vision method detects cigarette impurities, it is greatly affected by the shadow of the tobacco gap, and the color of the paper scraps i...

AI Technical Summary

Problems solved by technology

[0004] The invention provides a method for detecting sundries in shredded tobacco based on the Yolo V3 model, which solves the problems of low detection efficiency and low accuracy in the detection of sundries in shredded tobacco, improves the accuracy and work efficiency of the detection of sundries in shredded tobacco, and improves Cigarette production quality

Images