基于量子空间的粒子群算法在图像分割中的应用研究

RESEARCH ON THE APPLICATION OF QUANTUM SPACE BASED PARTICLE SWARM ALGORITHM TO IMAGE SEGMENTATION

研究图像的空间信息和灰度的图像分割,从中提取感兴趣的目标。传统的粒子群算法后期容易陷入早熟收敛状态,阈值选取时不能保证概率为1时收敛到全局最优解,导致计算时间延长,计算量增大,运算效率较低,抗噪能力差,最终造成分割效果不好。为了提高图像分割效率和分割精度,提出一种基于量子空间的粒子群算法的图像分割算法(QDPSO算法)。该方法通过最优阈值来划分像素,实现图像分割。实验结果表明,与传统的粒子群分割算法相比,该算法不仅得到了更高的分割精度,还大大减少了计算量,能够一定程度上改善图片分割的效率和质量。

By studying an image’s spatial information and grayed image segmentation,interesting targets are extracted from inside.Conventional particle swarm algorithms tend to fall into early mature convergence status.When selecting a threshold value,they can’t promise to converge toward a global optimal resolution at the probability rate of 1.Consequently their calculating time is prolonged,calculating workload is increased,computing efficiency is comparatively low,and noise-resistance capability is weak,in the end producing poor segmentation performances.In order to improve image segmentation efficiency and precision,a quantum space based particle swarm algorithm,as an image segmentation algorithm,named QDPSO algorithm,is proposed.The algorithm takes advantage of the optimal threshold value to partition pixels to realize image segmentation.Experiment results illustrate that,compared with conventional particle swarm segmentation algorithms,the proposed algorithm not only obtains higher segmenting precision,but also significantly decreases calculating workload so as to improve to a certain extent the efficiency and quality of image segmentation.