多光谱融合前端图像实时降噪处理系统的研究
|
摘要
多光谱图像融合技术无论是在军用还是民用领域都有着强劲的发展势头和广泛的市场应用前景。微光与红外图像的融合是夜视技术领域最为朝阳的一项新技术。然而,作为图像融合技术中的关键技术问题之一的图像融合前端图像预处理技术一直未能得到比较理想的解决。到目前为止,市场上还没有出现针对图像融合前端降噪优化预处理的处理系统。文章就“多光谱融合前端图像实时降噪处理系统的研究”这一课题,研制了一款小型化的、性能稳定的、针对微光与红外图像融合前端降噪预处理的实时图像处理器。
文章介绍了图像融合技术和图像融合预处理技术的基本发展;对图像的主客观评价方法进行了研究分析;针对微光和红外图像的各自特点提出了优化预处理方案;对各种图像优化处理的算法进行了分析和研究,特别针对图像的实时降噪进行了多模自适应递归降噪算法的研究;探索了适合于实时图像优化预处理的算法方案;完成了以FPGA为处理核心的小型化实时图像处理系统的整体解决方案;按照系统方案设计了硬件系统的各功能模块电路,并完成了小型化硬件系统的PCB设计;完成了基于FPGA的软件系统方案设计;以逻辑系统设计的方法为主并结合SOPC技术,完成了各软件功能模块的设计,成功搭建了图像预处理的算法实现平台;针对微光图像的预处理完成了多模自适应递归算法和灰度变换算法的软件设计;完成了整个软硬件系统的功能调试,并进行了实际环境中的图像处理实验和实验结果分析。
文章最终研制出了能够基本满足图像融合系统中前端图像实时降噪预处理的小型化处理器,并为更多功能和更有效的算法实现提供了成熟的平台。
The technology of multi-spectrum image fusion has a very good development tendency and a widespread market application prospect in both military and civilian domain. The fusion of the low level light and the infrared image is a very popular technology in the night vision domain. However, as one of the key technologies in the image fusion domain, the image pretreatment before the fusion still has no ideal solution. So far, there has not been any processing system to the optimal pretreatment of front-end image noise reduction in the market. The paper focus on the topic of "The research of the real-time image noise reduction system in front of the multi-spectrum fusion", and a small stable real-time image processor which is used to do noise reducing pretreatment in front of the fusion system of the low level light and the infrared image is developed.
In the paper, the development of the image fusion technology and pretreatment of the image fusion are introduced; Some image subjective and objective assessment methods are analyzed; The project of the pretreatment on the low level light and the infrared image are proposed respectively; Several algorithms to the optimal pretreatment of various kinds of images are analyzed, and the multimode auto-adapted recursion algorithm which is aimed at the image real-time noise reduction is studied specially; The project of the choice of algorithms for the real-time image optimal pretreatment is explored; The overall solution of the real-time image processing system which is based on FPGA is completed; According to the system project, various functional modules of the hardware system are designed, and the PCB design is completed; The project of the software system design based on the FPGA is completed; Various software functional modules are completed by the logical system design method Primarily and the SOPC technology, and the platform for the realizing of image pretreatment algorithms is built; The software designs of the multimode auto-adapted recursion algorithm and the gradation transformation algorithm which are chosen for the low level light image pretreatment are completed; The hardware and software system debugging, the test of image processing in the actual environments and the result analysis are completed.
Finally, a real-time noise reduction pretreatment processor which is able to satisfy the image fusion system basically and provide a mature platform for multi-purpose and the more effective algorithm realization is developed.
文章介绍了图像融合技术和图像融合预处理技术的基本发展;对图像的主客观评价方法进行了研究分析;针对微光和红外图像的各自特点提出了优化预处理方案;对各种图像优化处理的算法进行了分析和研究,特别针对图像的实时降噪进行了多模自适应递归降噪算法的研究;探索了适合于实时图像优化预处理的算法方案;完成了以FPGA为处理核心的小型化实时图像处理系统的整体解决方案;按照系统方案设计了硬件系统的各功能模块电路,并完成了小型化硬件系统的PCB设计;完成了基于FPGA的软件系统方案设计;以逻辑系统设计的方法为主并结合SOPC技术,完成了各软件功能模块的设计,成功搭建了图像预处理的算法实现平台;针对微光图像的预处理完成了多模自适应递归算法和灰度变换算法的软件设计;完成了整个软硬件系统的功能调试,并进行了实际环境中的图像处理实验和实验结果分析。
文章最终研制出了能够基本满足图像融合系统中前端图像实时降噪预处理的小型化处理器,并为更多功能和更有效的算法实现提供了成熟的平台。
The technology of multi-spectrum image fusion has a very good development tendency and a widespread market application prospect in both military and civilian domain. The fusion of the low level light and the infrared image is a very popular technology in the night vision domain. However, as one of the key technologies in the image fusion domain, the image pretreatment before the fusion still has no ideal solution. So far, there has not been any processing system to the optimal pretreatment of front-end image noise reduction in the market. The paper focus on the topic of "The research of the real-time image noise reduction system in front of the multi-spectrum fusion", and a small stable real-time image processor which is used to do noise reducing pretreatment in front of the fusion system of the low level light and the infrared image is developed.
In the paper, the development of the image fusion technology and pretreatment of the image fusion are introduced; Some image subjective and objective assessment methods are analyzed; The project of the pretreatment on the low level light and the infrared image are proposed respectively; Several algorithms to the optimal pretreatment of various kinds of images are analyzed, and the multimode auto-adapted recursion algorithm which is aimed at the image real-time noise reduction is studied specially; The project of the choice of algorithms for the real-time image optimal pretreatment is explored; The overall solution of the real-time image processing system which is based on FPGA is completed; According to the system project, various functional modules of the hardware system are designed, and the PCB design is completed; The project of the software system design based on the FPGA is completed; Various software functional modules are completed by the logical system design method Primarily and the SOPC technology, and the platform for the realizing of image pretreatment algorithms is built; The software designs of the multimode auto-adapted recursion algorithm and the gradation transformation algorithm which are chosen for the low level light image pretreatment are completed; The hardware and software system debugging, the test of image processing in the actual environments and the result analysis are completed.
Finally, a real-time noise reduction pretreatment processor which is able to satisfy the image fusion system basically and provide a mature platform for multi-purpose and the more effective algorithm realization is developed.
引文
[1]张敬贤,李玉丹,金伟其.微光与红外成像技术.北京:北京理工大学出版社.1995
[2]周立伟.微光成像技术的发展与展望.天津:天津科学技术出版社.2003
[3]周立伟.夜视技术的进展与展望.激光与光电子学进展.1995,30(4):37～43
[4]徐江涛,张兴社.微光像增强器的最新发展动向.应用光学.2005,26(2):21～23
[5]艾克聪.微光夜视技术的进展与展望.应用光学.2006,27(4):303～307
[6]刘卫光,李跃,张修社,范文新.图像信息融合与识别.北京:电子工业出版社.2008
[7]郭雷,李晖晖,鲍永生.图像融合.北京:电子工业出版社.2008
[8]曹倩.红外与微光图像融合实时处理器改进与算法研究.南京理工大学硕士学位论文.2007
[9]李卫华.数字图像预处理与融合方法研究.西北工业大学博士学位论文.2006
[10]D.L.Lampoa.Low Light and Thermal Imaging Systems.International Conference.1975
[11]刘咏.红外与微光图像融合实时处理系统硬件设计.南京理工大学硕士学位论文.2006
[12]王利平,刘滨,陈钱,张保民.多传感器目标识别的图像数据融合技术研究.光电子技术.1998,18(9):51～54
[13]王忆锋,张海联,李灿文,李茜.多传感器数据融合中的配准技术.红外与激光工程.1998,27(2):38～40
[14]韩崇昭,朱洪艳,段战胜.多源信息融合.北京:清华大学出版社.2006
[15]徐鲁安,叶懋冬,章琦.一种新的图像质量评价方法.计算机工程与设计.2004,25(3):418～420
[16]周景超,戴汝为,肖柏华.图像质量评价研究综述.计算机科学.2008,35(7):1～4
[17]LI WEN-juan,DU Hai-hui,DAI Jing-min and CHEN Ying-hang.A new method to evaluate imaging quality of CCD cameras.OPTOELECTRON1CS LETTERS.2005,1(2):148～150
[18]丁续星,朱日宏,李建欣.一种基于人眼视觉特性的图像质量评价.中国图象图形学报.2004,9(2):190～192
[19]李连胜,陈晚华.基于matlab的数字图像质量评价.湖南科技学院学报.2005,26(5):176～177
[20]佟雨兵,张其善,祁云平.基于PSNR与SSIM联合的图像质量评价模型.Journal of Image and Graphics.2006,11(12):1758～1763
[21]徐春梅,李刚,胡文刚,张伟.红外图像评价质量研究.红外技术.2004,26(6):72～75
[22]朱颖.自适应夜视图像处理技术研究.南京理工大学硕士学位论文.2007
[23]李俊山,李旭辉.数字图像处理.北京:清华大学出版社.2007
[24]景晓军,周贤伟,付娅丽.图像处理技术及应用.北京:国防工业出版社.2005.
[25]程源.多帧叠加及中值滤波在微光图像降噪中的应用.Radio Communications Technology.1997,23(3):34～36
[26]冯鹏,魏彪,米德伶,潘英俊,于渝.基于时域递归滤波的动态数字图像降噪.重庆大学学报(自然科学版).2005,28(2):23～25
[27]沈宇键,何昕,郝志航.基于时域递归低通滤波的实时背景预测系统的设计.微电子学.2001,31(1):23～27
[28]沈宇键,何昕,郝志航.时域递归滤波算法中k系数确定方法的研究.系统工程与电子技术.2001,23(3):61～65
[29]Jong-Hun Kim,Gang-Wook Song,Seong-Dae Kim.Lowpass Temporal Filter Using Motion Adaptive Spatial Filtering and Its Systolic Realization.IEEE Transactions on Consumer Electronics.1992,38(3):452～459
[30]Eric Dubois,Shaker Sabri.Noise Reduction in Image Sequences Using Motion-Compensated Temporal Filtering.IEEE Transactions on Communications.1984,32(7):826～831
[31]Francesco Cocchia,Sergio Carrato,and Giovanni Ramponi.Design and Real-Time Implementation of a 3-D Rational Filter for Edge Preserving Smoothing.IEEE Transactions on Consumer Electronics.1997,43(4):1291～1300
[32]冯卓煜.X射线图像图像处理算法研究与硬件实现.南京理工大学硕士学位文.2007
[33]姚晓.医用数字化X光机实时图像处理器的研制.南京理工大学硕士学位文.2006
[34]孙恋君.微测辐射热计阵列性能及信号处理电路研究.南京理工大学博士学位论文.2008
[35]林洪文,姚作梁.基于减背景技术的运动目标检测方法研究.国防科技大学学报.2003,25(3):66～69
[36]Philips Semiconductors.Data sheet for SAA7111AHZ.Philips Semiconductors Data Sheets.1998
[37]ANALOG DEVICES.Data sheet for AD8041.ANALOG DEVICES Data Sheets.1995
[38]ANALOG DEVICES.Data sheet for AD9235.ANALOG DEVICES Data Sheets.2004
[39]徐光辉,程东旭,黄如.基于FPGA的嵌入式开发与应用.北京:电子工业出版社.2006
[40]王金明.数字系统设计与Verilog HDL.北京:电子工业出版社.2007
[41]侯伯亨,顾新.VHDL硬件描述语言与数字逻辑电路设计.西安:西安电子科技大学出版社.1997
[42]ALTERA.Implementing Multipliers in FPGA Devices.ALTERA Corporation.2004
[43]CYPRESS.Data sheet for CY7C1061AV33.CYPRESS Data Sheets.2003
[44]ANALOG DEVICES.Data sheet for ADV7123.ANALOG DEVICES Data Sheets.1998
[45]ANALOG DEVICES.Data sheet for AD8042.ANALOG DEVICES Data Sheets.2006
[46]LINEAR TECHNOLGY.Data sheet for LT1963A Series.LINEAR TECHNOLGY Data Sheets.2000
[47]王俞允.PADS PCB高速电路设计.北京:中国电力出版社.2008
[48]江思敏,姚鹏翼.PADS电路原理图和PCB设计.北京:机械工业出版社.2007
[49]王玮冰.PADS电路设计.北京:国防工业出版社.2002
[50]Mark I.Montrose.电磁兼容的印制电路板设计.北京:机械工业出版社.2008
[51]邹澎,周晓萍.电磁兼容原理、技术和应用.北京:清华大学出版社.2007
[52]葛振南.非制冷红外热成像系统的小型化研究.南京理工大学硕士学位论文.2008
[53]周立功单片机发展有限公司.I~2C总线规范.广州周立功单片机发展有限公司.2000
[54]俞斯乐.电视原理.北京:国防工业出版社.2005
[55]何辅云,张海燕.电视原理与数字视频技术.合肥:合肥工业大学出版社.2003
[56]赵坚勇.电视原理与接收技术.北京:国防工业出版社.2007
[2]周立伟.微光成像技术的发展与展望.天津:天津科学技术出版社.2003
[3]周立伟.夜视技术的进展与展望.激光与光电子学进展.1995,30(4):37～43
[4]徐江涛,张兴社.微光像增强器的最新发展动向.应用光学.2005,26(2):21～23
[5]艾克聪.微光夜视技术的进展与展望.应用光学.2006,27(4):303～307
[6]刘卫光,李跃,张修社,范文新.图像信息融合与识别.北京:电子工业出版社.2008
[7]郭雷,李晖晖,鲍永生.图像融合.北京:电子工业出版社.2008
[8]曹倩.红外与微光图像融合实时处理器改进与算法研究.南京理工大学硕士学位论文.2007
[9]李卫华.数字图像预处理与融合方法研究.西北工业大学博士学位论文.2006
[10]D.L.Lampoa.Low Light and Thermal Imaging Systems.International Conference.1975
[11]刘咏.红外与微光图像融合实时处理系统硬件设计.南京理工大学硕士学位论文.2006
[12]王利平,刘滨,陈钱,张保民.多传感器目标识别的图像数据融合技术研究.光电子技术.1998,18(9):51～54
[13]王忆锋,张海联,李灿文,李茜.多传感器数据融合中的配准技术.红外与激光工程.1998,27(2):38～40
[14]韩崇昭,朱洪艳,段战胜.多源信息融合.北京:清华大学出版社.2006
[15]徐鲁安,叶懋冬,章琦.一种新的图像质量评价方法.计算机工程与设计.2004,25(3):418～420
[16]周景超,戴汝为,肖柏华.图像质量评价研究综述.计算机科学.2008,35(7):1～4
[17]LI WEN-juan,DU Hai-hui,DAI Jing-min and CHEN Ying-hang.A new method to evaluate imaging quality of CCD cameras.OPTOELECTRON1CS LETTERS.2005,1(2):148～150
[18]丁续星,朱日宏,李建欣.一种基于人眼视觉特性的图像质量评价.中国图象图形学报.2004,9(2):190～192
[19]李连胜,陈晚华.基于matlab的数字图像质量评价.湖南科技学院学报.2005,26(5):176～177
[20]佟雨兵,张其善,祁云平.基于PSNR与SSIM联合的图像质量评价模型.Journal of Image and Graphics.2006,11(12):1758～1763
[21]徐春梅,李刚,胡文刚,张伟.红外图像评价质量研究.红外技术.2004,26(6):72～75
[22]朱颖.自适应夜视图像处理技术研究.南京理工大学硕士学位论文.2007
[23]李俊山,李旭辉.数字图像处理.北京:清华大学出版社.2007
[24]景晓军,周贤伟,付娅丽.图像处理技术及应用.北京:国防工业出版社.2005.
[25]程源.多帧叠加及中值滤波在微光图像降噪中的应用.Radio Communications Technology.1997,23(3):34～36
[26]冯鹏,魏彪,米德伶,潘英俊,于渝.基于时域递归滤波的动态数字图像降噪.重庆大学学报(自然科学版).2005,28(2):23～25
[27]沈宇键,何昕,郝志航.基于时域递归低通滤波的实时背景预测系统的设计.微电子学.2001,31(1):23～27
[28]沈宇键,何昕,郝志航.时域递归滤波算法中k系数确定方法的研究.系统工程与电子技术.2001,23(3):61～65
[29]Jong-Hun Kim,Gang-Wook Song,Seong-Dae Kim.Lowpass Temporal Filter Using Motion Adaptive Spatial Filtering and Its Systolic Realization.IEEE Transactions on Consumer Electronics.1992,38(3):452～459
[30]Eric Dubois,Shaker Sabri.Noise Reduction in Image Sequences Using Motion-Compensated Temporal Filtering.IEEE Transactions on Communications.1984,32(7):826～831
[31]Francesco Cocchia,Sergio Carrato,and Giovanni Ramponi.Design and Real-Time Implementation of a 3-D Rational Filter for Edge Preserving Smoothing.IEEE Transactions on Consumer Electronics.1997,43(4):1291～1300
[32]冯卓煜.X射线图像图像处理算法研究与硬件实现.南京理工大学硕士学位文.2007
[33]姚晓.医用数字化X光机实时图像处理器的研制.南京理工大学硕士学位文.2006
[34]孙恋君.微测辐射热计阵列性能及信号处理电路研究.南京理工大学博士学位论文.2008
[35]林洪文,姚作梁.基于减背景技术的运动目标检测方法研究.国防科技大学学报.2003,25(3):66～69
[36]Philips Semiconductors.Data sheet for SAA7111AHZ.Philips Semiconductors Data Sheets.1998
[37]ANALOG DEVICES.Data sheet for AD8041.ANALOG DEVICES Data Sheets.1995
[38]ANALOG DEVICES.Data sheet for AD9235.ANALOG DEVICES Data Sheets.2004
[39]徐光辉,程东旭,黄如.基于FPGA的嵌入式开发与应用.北京:电子工业出版社.2006
[40]王金明.数字系统设计与Verilog HDL.北京:电子工业出版社.2007
[41]侯伯亨,顾新.VHDL硬件描述语言与数字逻辑电路设计.西安:西安电子科技大学出版社.1997
[42]ALTERA.Implementing Multipliers in FPGA Devices.ALTERA Corporation.2004
[43]CYPRESS.Data sheet for CY7C1061AV33.CYPRESS Data Sheets.2003
[44]ANALOG DEVICES.Data sheet for ADV7123.ANALOG DEVICES Data Sheets.1998
[45]ANALOG DEVICES.Data sheet for AD8042.ANALOG DEVICES Data Sheets.2006
[46]LINEAR TECHNOLGY.Data sheet for LT1963A Series.LINEAR TECHNOLGY Data Sheets.2000
[47]王俞允.PADS PCB高速电路设计.北京:中国电力出版社.2008
[48]江思敏,姚鹏翼.PADS电路原理图和PCB设计.北京:机械工业出版社.2007
[49]王玮冰.PADS电路设计.北京:国防工业出版社.2002
[50]Mark I.Montrose.电磁兼容的印制电路板设计.北京:机械工业出版社.2008
[51]邹澎,周晓萍.电磁兼容原理、技术和应用.北京:清华大学出版社.2007
[52]葛振南.非制冷红外热成像系统的小型化研究.南京理工大学硕士学位论文.2008
[53]周立功单片机发展有限公司.I~2C总线规范.广州周立功单片机发展有限公司.2000
[54]俞斯乐.电视原理.北京:国防工业出版社.2005
[55]何辅云,张海燕.电视原理与数字视频技术.合肥:合肥工业大学出版社.2003
[56]赵坚勇.电视原理与接收技术.北京:国防工业出版社.2007