A multi-label barcode fast recognition and positioning method and system in a warehouse environment

Abstract

The present application relates to the technical field of visual positioning, in particular to a multi-label barcode fast recognition and positioning method and system in a warehouse environment, comprising the following steps: collecting warehouse logistics images to construct a single-channel brightness feature matrix, calculating gradient vectors to construct a structure tensor field and solving texture coherence measure, generating a saliency probability distribution map and using an adaptive threshold to generate a high-frequency texture binary mask, performing a directional morphological closing operation according to the main texture flow direction to generate a closed connected region, extracting geometric moment features to calculate the centroid and output the position index. In the present application, the anisotropic barcode texture features are enhanced by constructing a structure tensor field, and the dynamic morphological closing operation logic based on the main texture flow direction effectively fills the texture gaps while suppressing background discrete noise, solving the positioning failure problem caused by feature aliasing in complex lighting and high-density multi-label scenes.

Images