医学图像配准算法研究

Multimodality Image Registration

研究图像配准精度优化问题,医学图像由多种图像结合,利用图像各自的特点进行融合。图像配准技术一直被广泛应用在医学图像和遥感图像领域中,针对传统的图像配准算法效率和精度较低等不足,为了提高医学图像配准的准确度,提出了一种将改进的最大熵算法并应用到图像配准的优化过程中,算法首先将输入的待配准图像进行灰度处理,对灰度值进行初始化,然后采用遗传算法的选择、交叉和变异操作对图像进行平滑,并选择最优值,最后采用最大熵算法对图像进行配准选择,算法有效克服了传统遗传算法容易陷入局部最优的缺点。仿真结果表明了改进的算法有效的提高了图像配准的精确度,验证了改进算法是有效的图像配准方法。

The problem of image registration accuracy. Image registration techniques has been widely used in medical imaging and remote sensing images, and other fields, for the traditional image registration algorithm and low efficiency and lack of precision, a genetic algorithm to improve adaptive and applied to image registration the optimi- zation process, the algorithm first pre-and post were adjusted using evolutionary crossover probability and mutation probability, the second cross, and immigration strategies to overcome the traditional genetic algorithm is easy to fall into local optimum shortcomings. Simulation results show that the algorithm effectively improves the accuracy of image registration, and with the common image registration algorithm compared to verify the feasibility of the method is an efficient image registration algorithm.