基于局部二次曲面拟合的SFS方法将图像中的每个像素视为具有决策能力的生命单元,以灰度图像为模板,在此模板引导下,模仿生物群体的自组织行为,实现自由曲面的三维重构。在图像平面的局部区域内,利用局部灰度信息拟合二次曲面来得到自由曲面的微观模型。微观模型通过自组织系统的计算,即可得到宏观的自由曲面模型。实验表明该方法能够快速得到理想的自由曲面,且比同类方法在计算精度上有显著提高。
In based on quadiie surface method, every pixel of the intensity image was regarded as a bion which had decisionmaking capacity, and the intensity image as target template. Guided by the target template, all pixels simulate the self-organization behavior of biologic colony to reconstruct the 3D surface. Micro models of the free-form surface were obtained by local shape recovery with quadratic approximation in the local area of the intensity image. The macro surface can be obtained by computation of self-organized system with all the micro models. The experiment results demonstrate this approach can reconstruct an ideal free-form surface, and has better precision than the traditional ones.