改进的基于神经网络的非均匀性校正算法

陈宝国; 张学峰; 牛英宇

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

改进的基于神经网络的非均匀性校正算法

陈宝国, 张学峰, 牛英宇

文章导航 > 红外与激光工程 > 2013 > 42(3): 574-578

陈宝国, 张学峰, 牛英宇. 改进的基于神经网络的非均匀性校正算法[J]. 红外与激光工程, 2013, 42(3): 574-578.

引用本文:

陈宝国, 张学峰, 牛英宇. 改进的基于神经网络的非均匀性校正算法[J]. 红外与激光工程, 2013, 42(3): 574-578.

Chen Baoguo, Zhang Xuefeng, Niu Yingyu. Improved nonuniformity correction algorithm based on neural network[J]. Infrared and Laser Engineering, 2013, 42(3): 574-578.

Citation:

Chen Baoguo, Zhang Xuefeng, Niu Yingyu. Improved nonuniformity correction algorithm based on neural network[J]. Infrared and Laser Engineering, 2013, 42(3): 574-578.

改进的基于神经网络的非均匀性校正算法

1.
西北工业大学电子信息学院,陕西西安710072;
2.
中国空空导弹研究院,河南洛阳471009

基金项目:

装备预先研究项目（61901020601）

详细信息

作者简介:
陈宝国（1970- ），男，研究员，主要从事红外探测系统方面的研究。Email：chenbg@163.com

中图分类号: TN216

Improved nonuniformity correction algorithm based on neural network

1.
School of Electronics and Information,Northwestern Polytechnical University,Xian 710072,China;
2.
China Airborne Missile Academy,Luoyang 471009,China

摘要: 提出了一种结合图像匹配和神经网络算法的焦平面阵列非均匀性校正算法。算法首先用最新的校正系数对图像进行非均匀性校正，输出校正结果；然后对相邻两帧图像进行匹配，估计出相邻帧之间图像的运动量；最后用神经网络算法分别对校正系数进行正向和反向自适应更新。采用图像匹配技术保证了校正系数更新时不会引起场景的模糊，采用校正系数双向更新策略可以保证每帧都能对每个像元的系数至少进行一次更新，与常用的神经网络校正算法相比，降低了对场景统计特性的要求，收敛速度较快。使用模拟添加噪声和采集的红外图像序列对算法进行仿真验证，结果表明，给出的算法校正效果优于常用的神经网络非均匀性校正算法。
- 非均匀性校正 /
- 神经网络 /
- 图像匹配 /
- 固定模式噪声 /
- 红外焦平面阵列
Abstract: An improved nonuniformity correction (NUC) algorithm combining image matching and neural network(NN) for infrared focal plane array sensors was presented. Firstly, nonuniformity of the FPA response was removed by NUC compensation. Then, motion parameters of the image were estimated by matching pairs of image frames. Finally, coefficients were adaptively updated according to bidirectional-renew strategy based on neural network. Image matching technique could effectively avoid faintness when coefficients were updating. Additionally, the bidirectional-renew strategy was used to guarantee coefficients of each pixel be calculated at least once when new image frame came. The new algorithm used image matching technique to get scene motion information, and used neural network for coefficients bidirectional-renew strategy. It had a lower statistical overhead on scenes and approached convergence more quickly than the often used neural network based NUC algorithms. A theoretical analysis was performed on a collection of infrared image frames to study the accuracy of the new NUC algorithm. It proves that it has higher-quality correction ability than simple neural network based NUC algorithm.
- nonuniformity correction /
- neural network /
- image matching /
- fixed-pattern noise /
- infrared focal plane array

[1]
[2]	Li Qing. Study on adaptive nonuniformity correction algorithm for IRFPA based on imaging guidance condition[D]. Xi'an: Xidian University, 2006. (in Chinese) 李庆. 基于成像制导状态的自适应IRFPA 非均匀性校正技术研究[D]. 西安: 西安电子科技大学, 2006.
[3]	Torres Sergio N, Hayat Majeed M. Kalman filtering for adaptive nonuniformity correction in infrared focal-plane arrays [J]. Optical Society of America, 2003, 20 (3): 470 -480.
[4]
[5]	Snyder D L, Angelisanti D, Smith W H, et al. Correction for nonuniformity flat-filed response in focal plane arrays[C]//SPIE, 1996, 2827: 60-67.
[6]
[7]	Ratliff Bradley M, Hayat Majeed M, Scott Tyo J. Generalized algebraic scene-based nonuniformity [J]. Optical Society of America, 2005, 22(2): 239-249.
[8]
[9]
[10]	Torres Sergio N, Vera Esteban M, Reeves Rodrigo A, et al. Adaptive scene-based non-uniformity correction method for infrared-focal plane arrays[C]//SPIE, 2003, 5076: 130-139.
[11]	Li Qing, Liu Shangqian, Wang Bingjian, et al. New nonuniformity correction algorithm for IRFPA based on neural network [J]. Infrared and Laser Engineering, 2007, 36(3): 342-344. (in Chinese) 李庆, 刘上乾, 王炳健, 等. 基于神经网络的IRFPA 非均匀性校正新算法[J]. 红外与激光工程, 2007, 36(3): 342-344.
[12]
[13]	Cai Shengbing, Duan Zhemin, Xu Jiadong. Scene-based nonuniformity correction method incorporating motion triggering[J]. Infrared and Laser Engineering, 2007, 36(6): 941-944. (in Chinese) 蔡胜兵, 段哲民, 许家栋. 加入运动触发场景的非均匀性校正方法[J]. 红外与激光工程, 2007, 36(6): 941-944.
[14]
[15]	Zhang Hao, Liu Zhenguo, Hu XiaoMei, et al. Nonuniformity correction based on active movement control of scene [J]. Infrared and Laser Engineering, 2011, 40(3): 397-401. (in Chinese) 张昊, 刘振国, 胡晓梅, 等. 场景主动运动控制的非均匀性校正方法[J]. 红外与激光工程, 2011, 40(3): 397-401.

[1]	汪大宝, 王中果, 吴明轩, 张超, 胡月. 基于改进伽马曲线的星载长波红外焦平面非均匀性校正算法 . 红外与激光工程, 2022, 51(5): 20210510-1-20210510-9. doi: 10.3788/IRLA20210510
[2]	陈文艺, 许洁, 杨辉. 利用双神经网络的相机标定方法 . 红外与激光工程, 2021, 50(11): 20210071-1-20210071-9. doi: 10.3788/IRLA20210071
[3]	沈英, 黄春红, 黄峰, 李杰, 朱梦娇, 王舒. 红外与可见光图像融合技术的研究进展 . 红外与激光工程, 2021, 50(9): 20200467-1-20200467-18. doi: 10.3788/IRLA20200467
[4]	张长兴, 刘成玉, 亓洪兴, 张东, 蔡能斌. 热红外高光谱成像仪光谱匹配盲元检测算法 . 红外与激光工程, 2020, 49(1): 0104002-0104002(7). doi: 10.3788/IRLA202049.0104002
[5]	尚超, 王锦春, 张晓兵. 红外焦平面阵列读出电路非均匀性研究 . 红外与激光工程, 2020, 49(8): 20190581-1-20190581-8. doi: 10.3788/IRLA20190581
[6]	钱润达, 赵东, 周慧鑫, 于君娜, 王士成, 荣生辉. 基于加权引导滤波与时域高通滤波的非均匀性校正算法 . 红外与激光工程, 2018, 47(12): 1204001-1204001(6). doi: 10.3788/IRLA201847.1204001
[7]	王天宇, 董文博, 王震宇. 基于单目视觉和固定靶标的位姿测量系统 . 红外与激光工程, 2017, 46(4): 427003-0427003(8). doi: 10.3788/IRLA201746.0427003
[8]	彭勇, 冯斌, 史泽林, 徐保树, 惠斌. 微偏振片阵列成像的非均匀校正研究 . 红外与激光工程, 2017, 46(4): 404004-0404004(8). doi: 10.3788/IRLA201746.0404004
[9]	王明昌, 樊养余, 陈宝国, 雷伟, 周波. 基于SOPC的红外图像自适应非均匀性校正设计 . 红外与激光工程, 2017, 46(6): 628001-0628001(6). doi: 10.3788/IRLA201746.0628001
[10]	郑博, 王健, 许文斌, 王晓东, 孙强. 基于方向性模型的红外焦平面阵列混叠效应研究 . 红外与激光工程, 2016, 45(11): 1104001-1104001(6). doi: 10.3788/IRLA201645.1104001
[11]	张磊, 王岳环, 宋琼. 红外焦平面条带状非均匀性校正硬件实现方法 . 红外与激光工程, 2015, 44(10): 2886-2890.
[12]	聂瑞杰, 李丽娟, 王朝林, 徐琰珂. 一种凝视红外成像系统联合非均匀校正算法 . 红外与激光工程, 2015, 44(8): 2339-2346.
[13]	许轰烈, 陈钱, 隋修宝, 刘宁. 基于单调场景运动的复杂非均匀性校正 . 红外与激光工程, 2014, 43(9): 3168-3172.
[14]	季尔优, 顾国华, 柏连发, 陈钱, 钱惟贤. 前景重配准的改进帧间误差最小化非均匀性校正算法 . 红外与激光工程, 2014, 43(5): 1672-1678.
[15]	刘慧, 朱明超, 吴泽鹏, 郭玲玲, 赵其昌, 贾宏光, 宣明. 红外焦平面阵列非均匀性自适应校正方法 . 红外与激光工程, 2014, 43(5): 1364-1369.
[16]	. 基于小波变换直方图规定的非均匀性校正算法 . 红外与激光工程, 2013, 42(12): 3481-3485.
[17]	任建乐, 陈钱, 顾国华, 钱惟贤. 红外焦平面阵列条纹非均匀性校正方法 . 红外与激光工程, 2013, 42(8): 1987-1990.
[18]	朱瑞飞, 王超, 魏群, 贾宏光, 周文明. 红外探测器非均匀性校正系统研制 . 红外与激光工程, 2013, 42(7): 1669-1673.
[19]	赵春晖, 刘振龙. 改进的红外图像神经网络非均匀性校正算法 . 红外与激光工程, 2013, 42(4): 1079-1083.
[20]	屈惠明, 陈钱. 环境温度补偿的红外焦平面阵列非均匀性校正 . 红外与激光工程, 2011, 40(12): 2328-2332.

点击查看大图

计量

文章访问数: 422
HTML全文浏览量: 68
PDF下载量: 184
被引次数: 0

改进的基于神经网络的非均匀性校正算法

1. 西北工业大学电子信息学院,陕西西安710072;

2. 中国空空导弹研究院,河南洛阳471009

作者简介:
陈宝国（1970- ），男，研究员，主要从事红外探测系统方面的研究。Email：chenbg@163.com

收稿日期: 2012-07-05
修回日期: 2012-08-13
刊出日期: 2013-03-25

基金项目:

装备预先研究项目（61901020601）

中图分类号: TN216

关键词:

摘要: 提出了一种结合图像匹配和神经网络算法的焦平面阵列非均匀性校正算法。算法首先用最新的校正系数对图像进行非均匀性校正，输出校正结果；然后对相邻两帧图像进行匹配，估计出相邻帧之间图像的运动量；最后用神经网络算法分别对校正系数进行正向和反向自适应更新。采用图像匹配技术保证了校正系数更新时不会引起场景的模糊，采用校正系数双向更新策略可以保证每帧都能对每个像元的系数至少进行一次更新，与常用的神经网络校正算法相比，降低了对场景统计特性的要求，收敛速度较快。使用模拟添加噪声和采集的红外图像序列对算法进行仿真验证，结果表明，给出的算法校正效果优于常用的神经网络非均匀性校正算法。

English Abstract

Improved nonuniformity correction algorithm based on neural network

1. School of Electronics and Information,Northwestern Polytechnical University,Xian 710072,China;

2. China Airborne Missile Academy,Luoyang 471009,China

Received Date: 2012-07-05
Rev Recd Date: 2012-08-13
Publish Date: 2013-03-25

Keywords:

Abstract: An improved nonuniformity correction (NUC) algorithm combining image matching and neural network(NN) for infrared focal plane array sensors was presented. Firstly, nonuniformity of the FPA response was removed by NUC compensation. Then, motion parameters of the image were estimated by matching pairs of image frames. Finally, coefficients were adaptively updated according to bidirectional-renew strategy based on neural network. Image matching technique could effectively avoid faintness when coefficients were updating. Additionally, the bidirectional-renew strategy was used to guarantee coefficients of each pixel be calculated at least once when new image frame came. The new algorithm used image matching technique to get scene motion information, and used neural network for coefficients bidirectional-renew strategy. It had a lower statistical overhead on scenes and approached convergence more quickly than the often used neural network based NUC algorithms. A theoretical analysis was performed on a collection of infrared image frames to study the accuracy of the new NUC algorithm. It proves that it has higher-quality correction ability than simple neural network based NUC algorithm.