Adaptive dimming belt foreign matter detection method and system based on fuzzy PID

Abstract

The invention discloses an adaptive dimming belt foreign matter detection method and system based on fuzzy PID. The system comprises an LED three-color illuminance module, a fuzzy PID adjustment illuminance module, an image gray processing module, an image acquisition module, a foreign matter detection hardware module, a foreign matter deep learning detection model module and a segmentation and sorting processing module. In consideration of the influence of a light source on image definition during image acquisition and a complex downhole environment, the illuminance of the light source is adjusted through fuzzy PID, the image definition is adjusted, a high-definition image input is ensured, the acquired image is input into a yolo-based deep learning detection model, and the recognized foreign matter image is segmented and recognized so that the precision of belt foreign matter recognition is effectively improved. By using the technical means of segmenting the image, the surface area of the belt foreign matter is calculated, so that the automation level of belt foreign matter sorting and fully mechanized caving face is improved, the labor intensity of workers is reduced, and the foreign matter sorting accuracy is ensured.

Description

technical field [0001] The invention relates to a belt foreign matter detection method and system, in particular to a fuzzy PID-based self-adaptive dimming belt foreign matter detection method and system, belonging to the technical field of image processing. Background technique [0002] Image recognition technology has been applied to the research of foreign object recognition in the field of belt transportation coal mines, but there is a problem of poor quality of collected images, which leads to low recognition accuracy. [0003] The current research is mainly aimed at the sorting of foreign matter in coal transportation, identifying the flat foreign matter on the belt. Due to the large number of types, such as discarded anchors, bolts and wood, etc., it is more difficult to sort foreign matter. Moreover, the environment for collecting images is poor, and the quality of input images cannot be guaranteed. And in the process of image acquisition, usually a linear light sou...

AI Technical Summary

Problems solved by technology

[0002] Image recognition technology has been applied to the research of foreign object recognition in the field of belt transportation coal mines, but there is a problem of poor quality of collected images, which leads to low recognition accuracy

[0003] The current research is mainly aimed at the sorting of foreign matter in coal transportation, identifying the flat foreign matter on the belt. Due to the large number of types, such as discarded anchors, bolts and wood, etc., it is more difficult to sort foreign matter. Moreover, the environment for collecting images is poor, and the quality of input images cannot be guaranteed

And in the process of image acquisition, usually without considering the illuminance factor, a linear light source is set arbitrarily, or several different light sources are set, but there is no dimming control for the light illuminance

In fact, different foreign objects determine that their response characteristics to light are also different, that is, under different illumination, even the same object will be visually different, so the current image-based belt foreign object recognition technology has the problem of poor accuracy

Images