Comparative Analysis of Segmentation Efficiency Method the Halftont Image Based on the Selection of Priority Direction of Machining Segment Boundaries

摘要

This article proposes a method for segmenting raster halftone images enhances the quality and accuracy of the allocation of the image. The developed method is based on post gradient image processing, whereby a sliding nine elements window selects a priority processing of images and boundaries are corrected pixel brightness gradient image, caught in the priority areas of the sliding window. Implemented analysis of the effectiveness of the proposed method in comparison with one of its most well-known, the gradient method. An assessment of the complexity of the algorithms implemented by comparing the number of operations performed in the present and the known method. The graph of the time implementing a gradient algorithm and algorithm selection processing of the priority areas is presented, taking into account the size of the processed image. It offers a variety of options to determine the accuracy and speed of segmentation algorithms. The estimation accuracy for the three types of segmentation algorithms and built for the allocation of the relevant characteristics of circuits.