摘要
针对大多数图像分割方法计算量大、不利于实时处理的缺点,提出用微粒群算法(PSO)优化最小误差分割方法。该方法不但具备最小误差分割法受目标和噪声影响小以及对小图像分割效果好的优点,还克服了遗传算法等加速算法需要预先设定众多运行参数,受目标变化影响大的问题。图像分割的效果和速度得到了提高,性能也更加稳定。实验结果反映了该方法的有效性。
Owing to the defects of which most image segmentation method's calculation is burdensome and not fit for a real-time system, an approach using minimum error method based on particle swarm optimization is proposed. This method not only has a good result for small targets with more yawp, but also solue the problem of which genetic algorithm and other accelerate algorithm needed enacted many parameters and affected by target change. The effects and speed of image segmentation have been improved and the performance become more steady. Experimental results show the feasibility of the proposed methods.
出处
《计算机工程与设计》
CSCD
北大核心
2006年第18期3377-3378,3387,共3页
Computer Engineering and Design
基金
浙江省教育厅基金项目(20031165)。
关键词
微粒群算法
最小误差
图像分割
遗传算法
实时处理
particle swarm optimization
minimum error
image segmentation
genetic algorithm
real-time system
real-time process