一种混合遗传算法在图像分割中的应用

Image Segmentation Using Hybrid Genetic Algorithm

在图像分割中,最小误差法计算简单,受目标和噪声影响小,对小图像仍具很好的分割效果,但计算量大,不利于实时处理。为解决这一问题,该文将遗传算法和退火算法引入到最小误差法中,结合遗传算法的全局寻优能力和模拟退火算法较强的局部搜索能力,提出一种高效的混合遗传算法(GASA),充分利用该混合算法快速和稳定性强的优点来减少最小误差法的运算量,不仅能够提高运算收敛速度和收敛效率,而且可以有效避免出现早熟现象,防止陷入局部最优,同时性能也很稳定,完全能满足实时系统中精度和速度的要求,得到较好的分割效果。

In the imaging segmentation, the Minimum Error is a better method because of its simplicity and having good result for small targets with more yawp. But it is not fit for a real-time system because its calculation is burdensome and will spend lots of time to segment images. In solving this problem an effective hybrid genetic algorithm (GASA) is proposed which combines the capacity of GA to reach the global optimum with the capability of SA to gain the local one. This text makes use of the fleetness and stability of GASA to reduce the calculation. GASA can not only enhance the rate and efficiency of algorithmic constringency but also effectively avoid appearing precocity and plunging into local optimum. This method with steady performance can completely satisfy the accuracy and speed′s requirements of a real-time system and provide a better effect on image segmentation.