Multiscale Statistical Analysis of Massive Corrosion Pits Based on Image Recognition of High Resolution and Large Field-of-View Images

Abstract

In the present study, a new multiscale method is proposed for the statistical analysis of spatial distribution of massive corrosion pits, based on the image recognition of high resolution and large field-of-view (montage) optical images. Pitting corrosion for high strength pipeline steel exposed to sodium chloride solution was observed using an optical microscope. Montage images of the corrosion pits were obtained, with a single image containing a large number of corrosion pits. The diameters and locations of all the pits were determined simultaneously using an image recognition algorithm, followed by statistical analysis of the two-dimensional spatial point pattern. The multiscale spatial distributions of pits were analyzed by dividing the montage image into a number of different windows. The results indicate the clear dependence of distribution features on the spatial scales. The proposed method can provide a better understanding of the pit growth from the perspective of multiscale spatial evolution.

Rapid Classification of Massive Images Based on Cloud Computing Platform

“…Therefore, the traditional image processing mode has gradually failed to meet the needs of users. Now it is particularly necessary to seek a new and efficient image processing mode [13][14][15][16][17][18][19][20]. As a new calculation pattern, cloud computing is a typical distributed and parallel computing model, which can greatly shorten the execution time of computing tasks.…”