【摘要】 现有图像清晰度评价函数在对ICF微靶装配体图像的评价过程中,由于不同零件图像对比度的变化,评价函数极易失去其理想曲线特性,导致聚焦精度降低甚至调焦过程失败。针对这一问题,在分析离焦图像模糊原因的基础上,得出离焦图像究其本质为图像结构变化失真。利用Zernike矩具有描述图像结构特征这一特性,提出一种新的针对微靶装配体图像的清晰度评价函数。评价函数由图像不同阶Zernike矩的线性组合构成,通过调节权重系数来实现微靶装配体不同零件图像的清晰度评价。实验结果表明:相较于现有的评价函数,新的评价函数在对微靶装配体图像评价过程中保持了函数的理想曲线特性,尤其是针对低对比度图像,其在灵敏度和抗噪性方面具有明显优势。更多还原

【Abstract】 During the process of evaluating ICF micro-target image with the current image definition evaluation function,owing to the changes in image contrast of different components of micro-target,the evaluation function may lose its ideal curve characteristics easily. This may not only result in the decrease of focus accuracy but also,in some cases,the failure of the entire focusing process. Based on the analysis of the causes for the blurring defocused image,we discovered that the primary reason for defocused image was structural distortion. With the characteristic of Zernike moment of describing the image structure in mind,we propose a new function for evaluating the image definition of micro-target. The new evaluation function is consisted of Zernike moments weighting of different steps,and it is realized through the evaluation of image contrast for micro-targets of different definitions by regulating the weight coefficient. We confirm that the new evaluation function maintains the ideal curve characteristics during the process of evaluating the micro-target images compared to the current evaluation function. In the condition of low contrast,however,the new approach holds a distinct advantage in terms of sensitivity and noise immunity.更多还原