Application of a New Segmentation Algorithm in Submarine Manganese Nodule Statistics

Qing-wen CAI, Yu-liang LIU

Abstract


To solve the problem of particle overlapping in the statistics of submarine manganese nodules, a new algorithm is proposed in this paper. According to the natural round nucleus shape of manganese nodule particle, a substitution circle is used to replace the particle approximately. Firstly, the centroid of particle obtained by calculating is regarded as the center of the substitution circle. Then, the distance between the circle’s center and the boundary of the particle is obtained, and the distance plus to a modified constant as the radius of the substitution circle. Finally, the intersection points of two substitution circles are looked upon as the pair of segmentation points, and the manganese nodule particles is segmented by connecting a line between the pair of segmentation points. From experiments results, it’s shown that the accuracy of statistics is over 98% after processed by the new algorithm, which proved the proposed algorithm is effective.

Keywords


Manganese nodule, Segmentation, Substitution circle, Splitting points pair, Centroid


DOI
10.12783/dtcse/ccme2018/28674

Refbacks

  • There are currently no refbacks.