用于全数字对焦的点扩散函数性能分析与评价

Performance analysis and evaluation of point spread function for full digital autofocus

为了对用于全数字对焦的点扩散函数模型进行定量分析,提出用原始图像与处理所得图像的均方差进行性能评价的方法.点扩散函数模型分别被用于全数字对焦处理,并通过基于点扩散函数的图像恢复,完成全数字对焦的功能模块和评价模块.结果显示,高斯型点扩散函数模型具有最佳的性能,在评价指标方面优于均匀扩散模型和基于刃边响应的点扩散函数模型.高斯模型能够恢复出更接近于准确对焦的图像,而且对被成像物体没有直边特征的限制.

To quantificationally analyze point spread function （PSF） models for full digital autofocus method, an evaluation method of mean square error between original image and processed image was proposed for comparison performance of algorithms. PSF models were introduced into full digital autofocus, and theirs function module and evaluation module were implemented using image restoration. The analysis results show that Gaussian PSF model has the best performance. The model can acquire better restoration result than circular disk PSF model and the PSF models based on edge response, and has not restrict of distinct edge for imaging objects.