A tank truck tank opening flange visual inspection method and system for low-illumination environments

Abstract

The present application relates to the technical field of automatic loading and unloading of petrochemical industry fluids, and discloses a tank truck tank opening flange visual detection method for low-illumination environments, comprising the following steps: S1: collecting image data of the tank opening flange of the oil tank truck under the interference of light, and dividing the collected image data into a training set, a verification set and a test set after labeling; S2: based on a YOLOv11n detection model, a low-illumination feature enhancement module is fused to construct a YOLO-LLFE detection model; wherein the low-illumination feature enhancement module is used to improve the detection accuracy of the model in a low-illumination scene. Through double pooling operations in the horizontal and vertical directions, the model can comprehensively capture the contour features of the tank opening flange in the spatial structure, even in the case of texture blur caused by low illumination, the structural expression ability can still be effectively enhanced, and the accuracy of feature extraction can be significantly improved.

Images