针对计算机辅助颅像重合技术中三维颅骨模型与二维人脸照片的颅像叠加问题,提出一种基于浮点数编码遗传算法的自动叠加方法。首先使用遗传算法寻找三维颅骨相对于二维照片上对应标志点的最佳匹配几何变换f,遗传算法采用浮点数编码,选择、交叉、变异算子分别使用最佳保留选择算子、启发式交叉算子和随机变异算子,然后通过相似性变换和透视投影变换,将三维颅骨模型映射到二维人脸照片上,实现最佳颅像叠加。实验表明,此方法实现了自动化的颅像叠加,且耗时短,准确度高。
Regarding the photo-skull overlay process in computer-aided photographic superimposition,this paper proposed an automated method based on a real-coded genetic algorithm. First,it used the genetic algorithm to search for the geometric transformation f,which matched 3D skull model with 2D photo optimally according to corresponding landmarks. The selection operator,crossover operator and mutation operator of the real-coded genetic algorithm are best retention operator,Wright's heuristic crossover and a random mutation operator respectively. Then it mapped the 3D skull model onto the 2D photo using a similarity transformation and a perspective projection transformation. The experiments show that this method realizes the process of skull-photo overlay automatically,fast and accurately.