Foreign matter detection method and system for tobacco rolling and packaging equipment

Abstract

The invention discloses a foreign matter detection method and system for tobacco rolling and packaging equipment, which is characterized in that a foreign matter detection model for the tobacco rolling and packaging equipment is established on the basis of a YOLOv3 algorithm, and training of training samples and testing and re-optimization of a test data set are combined, so that the accuracy of the foreign matter detection model for the tobacco rolling and packaging equipment is effectively improved, and the labor cost is greatly reduced; meanwhile, different threshold values are set for different types of foreign matters according to test results, so that the detection accuracy is further improved, and the problem of unstable model accuracy caused by imbalance of difficult samples is solved.

Description

technical field [0001] The invention belongs to the technical field of target detection, and in particular relates to a foreign object detection method and system for tobacco wrapping equipment. Background technique [0002] With the advancement of science and technology and the improvement of hardware computing power, the method of manual monitoring and review is far from meeting the needs. Therefore, the application of intelligent monitoring technology was born and quickly became a research hotspot. The maintenance and inspection of equipment in the packaging workshop of a cigarette factory is an important part of the equipment guarantee work in the workshop. The quality of equipment maintenance and inspection is related to the completion of the production target of the workshop, the guarantee of product quality and the control of consumption indicators. At present, the evaluation of equipment maintenance in the packing workshop adopts a hierarchical inspection mode. The s...

AI Technical Summary

Problems solved by technology

This method is mainly aimed at the detection of wet lump tobacco, and some foreign objects left after the maintenance of the equipment in the packing workshop, such as tobacco, tobacco dust, rubber dirt, etc., cannot be detected at fixed points

[0006] Therefore, at present, the cigarette factory mainly checks the maintenance of the equipment in the packing workshop during the handover period through manual auditing. However, the manual auditing method has the following problems: First, manual auditing involves many subjective factors. The value and other factors will affect the review results; the second is that it is difficult to distinguish the foreign matter on the equipment due to the aging of the equipment, the main reason is that the background complexity is high, the foreign matter is more complicated, and the foreign matter is more complicated. Judgment is greatly affected by subjective factors, and the size of foreign objects is uncertain. For example, for small shredded tobacco or the influence of light during night shift review, artificial vision is difficult to meet the needs of various occasions. Therefore, manual review not only consumes a lot of manpower, time and Administrative cost, and inefficiency, high rework rate

Images