节点文献
优良模式自学习遗传算法在阈值选取中的应用
The Excellent Schema Self-learning Genetic Algorithm Approach to the Selection of Image Threshold
【摘要】 针对现有图像处理过程中阈值选取优化方法中存在的计算效率低、易陷入局部最优等不足,研究采用优良模式自学习算法,求取图像阈值;通过对类间方差的优化,表明该文算法优越性。实验结果进一步证实了该文算法优良、高效。针对图像特点,利用优良模式自学习遗传算法,提出了一种阈值自动选取的策略,提高了阈值选取的准确性及寻优速度。实验结果表明了该文提出算法的可行性。
【Abstract】 In this paper, an optimal strategy of automatic selection of image threshold based on an excellent schema self-learning genetic algorithm is studied. Using it to maximum classes square error shows that this method can improve accuracy during selecting the thresholds and accelerate the optimal process. The experimental result also confirms that the new strategy is very effective on selecting image thresholds.
【关键词】 阈值;
类间方差;
熵判据;
遗传算法;
【Key words】 Threshold; Classes square error; Entropy criterion; Genetic algorithms;
【Key words】 Threshold; Classes square error; Entropy criterion; Genetic algorithms;
【基金】 国家自然科学基金资助项目(69874031)
- 【文献出处】 计算机工程 ,Computer Engineering , 编辑部邮箱 ,2002年09期
- 【分类号】TP391.4
- 【被引频次】5
- 【下载频次】102