Arsenide storage tank health state diagnosis method and system based on industrial vision

Abstract

The invention relates to the technical field of industrial safety monitoring, in particular to an arsenic hydride storage tank health state diagnosis method and system based on industrial vision. The method comprises the following steps: collecting multi-moment images through industrial vision in a stable operation stage of the storage tank, dividing a stable observation area through geometric alignment and structural constraint, extracting visual features reflecting the shape consistency of the tank body, and establishing a visual health reference through statistical analysis; images are continuously collected in the operation stage, multi-scale visual features are extracted, and a feature time sequence is constructed to describe structural state evolution; by calculating the offset of the operation characteristics and the health benchmark, an overall evolution offset index is obtained through smooth processing and weighted aggregation, and quantitative judgment of the health state is achieved in combination with sliding window trend analysis and dynamic threshold judgment; and performing short-term trend prediction based on the historical offset sequence, and generating prediction and early warning information through structural sensitivity correction and risk grade division. According to the invention, closed-loop health diagnosis from state monitoring to trend pre-judgment is realized.

Images