大数据分析的红外图像非均匀性自动校正系统

Infrared image nonuniformity automatic correction system based on big data analysis

现有的红外图像自动校正系统在计算像素点非均匀性位移时误差较大,导致系统在校正红外图像时精度较差,设计基于大数据分析的红外图像非均匀性自动校正系统。在系统硬件中,设计传感器接口模块,设计模数转换接口模块,保证系统可以在大量的数据中完成分析工作。在软件设计中,生成图像校正条件判别网络,基于大数据分析配准红外图像非均匀性参数,建立图像非均匀性自动校正模型,得到红外图像非均匀性自动校正系统。实验结果表明,在学习效率为0.015时,均方根误差为20%,在不同非均匀性倍数为2时,设计系统的校正精度为83%,因此,通过大数据分析方法可以得到更加清晰明确的红外图像,实现非均匀性自动校正。

The existing infrared image automatic correction system has a large error in calculating the pixel non-uniformity displacement,resulting in poor accuracy in correcting the infrared image.An infrared image non-uniformity automatic correction system based on big data analysis is designed.In the system hardware,the sensor interface module and analog-to-digital conversion interface module are designed to ensure that the system can complete the analysis in a large amount of data.In the software design,the image correction condition discrimination network is generated,the infrared image nonuniformity parameters are registered based on big data analysis,the image nonuniformity automatic correction model is established,and the infrared image nonuniformity automatic correction system is obtained.The experimental results show that when the learning efficiency is 0.015,the root mean square error is 20%,and when the different non-uniformity multiple is 2,the correction accuracy of the designed system is 83%.Therefore,a clearer infrared image can be obtained through the big data analysis method to realize the automatic correction of non-uniformity.