摘要
通过图像清晰度评价函数判断离焦图像的清晰度是相机实现自动调焦的关键。使用最优离散余弦变换(DCT)系数的和作为评价图像清晰度方法,并用该评价方法与图像的调制传递函数(MTF)的变化趋势做相关性分析,以验证该评价方法的有效性。结果表明:基于离散余弦变换的评价方法可以得到与MTF相似的评价结果,归一化后,最清晰图像的判断值都为1,随图像变模糊时函数的下降趋势基本一致。
Determine defocus image sharpness through image sharpness evaluation function is the key to realize camera auto-focusing,using sum of optimal discrete cosine transform （DCT）coefficients as method for evaluating image sharpness, and do correlation analysis using this method with the image modulation transfer function （MTF） ,in order to verify the effectiveness of evaluation method, the results show that evaluation method based on discrete cosine transform coincide with the MTF. After normalization, the best image values are 1, the function decrease with the image becomes hazing.
出处
《传感器与微系统》
CSCD
2017年第2期67-70,共4页
Transducer and Microsystem Technologies
关键词
图像清晰度
调制传递函数
离散余弦变换
相关性
image sharpness
modulation transfer function（MTF）
discrete cosine transform（DCT）
correlation
