一种新的清晰度评价算法  

New Definition Evaluation Algorithm

作　　者：余军[1] 王沛[1] 

机构地区：[1]上海师范大学信息与机电工程学院,上海200235

出　　处：《电视技术》2013年第11期75-78,共4页Video Engineering

摘　　要：高效、准确的图像清晰度评价是实现基于分块的多聚焦图像融合的关键。在分析了图像清晰度影响因素和现有图像清晰度评价方法优缺点的基础上,提出了一种改进的应用平均梯度函数的清晰度评价,建立了评价清晰度函数的数学模型,并对不同聚焦图像进行实验和仿真,比较了各种评价函数。结果表明,该方法具有独立性好、精确度高、对比度明显等优点,能够应用于实际的图像清晰度评价和多聚焦融合中,具有较高的实用性和可靠性。The efficient and accurate image definition evaluation is a key step in block-based multi-focus image fusion. After analyzing the advantages and disadvantages of existing methods for image definition evaluation and impact factors of image definition, in this paper, a new method is proposed based on the improved average gradient of definition evaluation for image definition evaluation. Then, the evaluation function of the mathematical model is built. Experiments and simulations are made with different focus image. It is proved that, this evaluation method has high independence and accuracy, and the contrast is obvious, has a high level of practicality and reliability in image clarity evaluation and multi-focus fusion.

关 键 词：清晰度评价 多聚焦图像融合 数学模型 

分 类 号：TN911.73[电子电信—通信与信息系统] TP941.1[电子电信—信息与通信工程]