Design and implementation of a parallel OTSU algorithm based on a ternary optical computer

Abstract

The OTSU method (OTSU) is considered to be the best algorithm for threshold selection in image segmentation. The algorithm is easy to calculate and not affected by image brightness and contrast, so it is widely used in the field of digital image processing. Due to the slow running speed and large storage space of the algorithm on the traditional electronic computer, this research makes use of the advantages of the ternary optical computer (TOC), such as numerous data bits and the reconfigurable processor. Through the analysis of the OTSU algorithm, we find the part of the OTSU algorithm that can be calculated in parallel, and put forward a parallel OTSU algorithm based on TOC. The algorithm first uses a classifier to classify the number of pixels. By reconstructing the MSD adder and multiplier, the inter-class variance under each gray value is calculated in parallel, and the maximum value is selected by comparing the inter-class variance. The corresponding threshold is the best segmentation threshold. By analyzing the clock cycle and data bit resources of the algorithm, and then conducting a comparative analysis and experimental verification, we found that the computational speed of TOC increased by approximately 12% when handling images with the same amount of computation.