基于PSO和模糊划分熵的水下图像分割

在水下环境中,由于存在着水体对光线的吸收以及照明不均等原因,水下图像具有信噪比低、边缘模糊等特点。如果直接使用传统的分割方法,对水下图像进行处理后的效果较差。传统的基于最大熵原理的阈值法尽管能实现某些特定的分割任务,但是其时效性较差。而粒子群算法(PSO)是一类随机全局优化技术,该算法简单易实现,可调参数少。因此将群体智能中的粒子群优化算法应用到图像分割中。新方法在重新定义模糊熵的基础上,根据最大熵原理,利用粒子群算法来搜索分割阈值。相对于传统的利用穷举法来搜索分割阈值的算法,新方法大大减少了计算时间,提高了效率。通过对水下图像处理实验证明,该算法对简单背景的图像分割是有效的,和传统分割方法相比,具有更强的自适应性和抗噪性能。

Underwater image segmentation based on particle swarm optimization and fuzzy partition entropy

Due to the assimilation of the water and uneven lightness,the underwater images would have low S/N and the detail is fuzzy.If traditional methods are used to dispose underwater images directly,it is unlikely to obtain satisfactory results.Though traditional threshold methods based on maximum entropy principle could sometimes divide the image into object and background,its time-consuming computation is often an obstacle.Particle swarm optimization(PSO) is a stochastic global optimization technique,which has become the hotspot of evolutionary computation because of its excellent performance and simple for implement.The particle swarm optimization algorithm is applied into the image segmentation.In the method,fuzzy entropy is redefined on given images,and the particle swarm optimization algorithm is used to search the optimal threshold based on maximum entropy principle.The experiments prove that this novel approach is effective for simple back-grounded underwater images.Comparing with the traditional methods,the new method shows better adaptability and noise restraining performance.