Combining Focus Measures through Genetic Algorithm for Shape from Focus

摘要: For the reconstruction of three-dimensional (3D) shape microscopic objects different focus measure operators have been employed. It is difficult to compute accurate depth map using a single due type texture. Moreover, real images with diverse types illumination and contrast lead erroneous estimation through measure. So address this problem, we used spatial in conjunction genetic algorithm for estimation. Genetic uses output weight updating method accurately estimates maps world objects. The performance developed then evaluated by both synthetic image sequences. experimental results show that proposed more useful computing as compared existing SFF methods.