基于清晰度探测与人机交互的图像质量评价算法

Image quality assessment system based on resolution detection and human-computer interaction

为了解决清晰度函数种类覆盖度小和人机交互缺乏灵活性的背景下,存在图像质量评价力不足的问题,分别从清晰度探测数学模型分析和人机交互软件开发的角度出发,提出了基于清晰度探测与人机交互的图像质量评价算法。首先,根据图像灰度标准差函数、熵函数和傅里叶变换函数,设计了耦合灰度空间域、能量域与频域的清晰度探测算子,多领域评价图像清晰程度。然后,根据Windows消息响应机制,推导图像与控件尺寸转换关系,结合图像部分区域提取方法,实现图像感兴趣区域提取目的。最后,在感兴趣图像区域进行清晰度评价算子计算,开发图像质量评价系统。实验测试结果显示:与当前图像质量评价机制相比,本文机制拥有更健壮的评价力与实用性。

In order to solve the current version of the official software is expensive and the definition of function types cov- ered by small and interactive lack of universality in the background, lack of image quality evaluation problems, this paper from the definition of detection model analysis and interactive software development perspective, put forward the image quality evaluation system based on man-machine interaction detection and resolution. Firstly, according to the image gray standard deviation function, the entropy function and the Fu Liye transform function; we design the detection operator of the coupling gray space domain, the energy domain and the frequency domain. Then, according to the Windows message response mechanism, the relationship between the image and the size of the control is deduced, and then the region of in- terest is extracted with the method of image region extraction. Finally, the image quality evaluation system is developed by integrating the clarity detection operator and the human-computer interaction mechanism. The experimental results show that compared with the current image quality evaluation mechanism, this mechanism has higher precision and prac- ticability.