基于改进遗传模拟退火算法的BP神经网络的畸变校正研究
|
摘要
针对相机所采集的图像大多都存在畸变现象的问题,设计了基于改进遗传模拟退火算法的BP神经网络校正算法;该算法针对传统遗传算法易于收敛局部最优的问题,提出分段选择策略与随机抽样相结合的选择算子,自适应交叉与变异算子;在畸变校正中,该算法通过网络的输入输出建立理想点与畸变点的关系,使用改进的遗传模拟退火算法来优化神经网络中的阈值与权值,然后使用基于LM算法的BP神经网络进行局部优化,最后通过插值算法得到校正后的图像;实验表明,该算法能够较好的对图像进行畸变校正,同时与传统的BP神经网络算法相比精度更高,收敛速度更快。
Aimed at the problem that the images collected by the camera are mostly distorted,a BP neural network correction algorithm based on improved genetic simulated annealing algorithm is designed.Aiming at the problem that traditional genetic algorithm is easy to converge local optimum,this algorithm proposes a selection operator combining segmentation selection strategy and random sampling,adaptive crossover operator and adaptive mutation operator.In the distortion correction,the algorithm establishes the relationship between the ideal point and the distortion point through the input and output of the network.Optimizing the threshold and weight in the neural network through improved genetic simulated annealing algorithm,then carring out local optimization through BP neural network based on LM algorithm,and finally obtaining the corrected images by interpolation.Experiments show that the proposed algorithm can correct the distortion of the image better than the traditional BP neural network algorithm,and the convergence speed is faster.
Aimed at the problem that the images collected by the camera are mostly distorted,a BP neural network correction algorithm based on improved genetic simulated annealing algorithm is designed.Aiming at the problem that traditional genetic algorithm is easy to converge local optimum,this algorithm proposes a selection operator combining segmentation selection strategy and random sampling,adaptive crossover operator and adaptive mutation operator.In the distortion correction,the algorithm establishes the relationship between the ideal point and the distortion point through the input and output of the network.Optimizing the threshold and weight in the neural network through improved genetic simulated annealing algorithm,then carring out local optimization through BP neural network based on LM algorithm,and finally obtaining the corrected images by interpolation.Experiments show that the proposed algorithm can correct the distortion of the image better than the traditional BP neural network algorithm,and the convergence speed is faster.
引文
[1]冯广.基于粒子群参数优化的多个自由平面标定法的图像畸变校正研究[D].昆明:昆明理工大学,2017.
[2]Tsai R Y.A versatile camera calibration technique for high accuracy SD machine vision metrology using off-the-shelf TVcameras and lenses[J].IEEE Jourmal of Robotics Automation.1987,3(4):323-344.
[3]Zhang Z.A flexible new technique for camera calibration[J].IEEE Transactions on Pattern Analysis and Machine Intelligence.2000,22(11):1330-1334.
[4]叶峰,王敏,陈剑东,等.共面点的摄像机非线性畸变校正[J].光学精密工程,2015,23(10):2962-2970.
[5]杨麒,李天伟,黄谦,等.基于直线投影特征的镜头畸变校正方法[J].计算机测量与控制,2018,26(9):264-268.
[6]Song G Y,Lee J W.Correction of radial distortion based on line-fitting[J].International Journal of Control,Automation and Systems,IEEE,2010,18(3):615-621.
[7]白福忠,王建新,杨慧珍,等.视觉测量技术基础[M].北京:电子工业出版社,2013.
[8]Weng J,Cohen P,Herniou M.Camera Calibration with Distortion Models and Accuracy Evaluation,IEEE Trans.Pattern A-nalysis and Machine Intelligence,1992,14(10):965-980.
[9]王红平,曹国华,李振辉,等.基于BP网络的图像畸变校正技术研究[J].长春理工大学学报(自然科学版),2009,32(4):653-656.
[10]江祥奎,范永青,王婉.基于粒子群遗传算法的BP神经网络摄像机标定[J].计算机科学与探索,2014,8(10):1254-1262.
[11]任贝,韩飞,吴坚.基于遗传算法的摄像机标定[J].吉林大学学报(信息科学版),2013,31(4):432-436.
[12]韩啸.基于遗传算法的摄像机内参数标定研究[D].长春:吉林大学,2008.
[13]徐正宏.基于遗传-模拟退火混合算法的泰兴市高压配电网优化规范研究[D].北京:华北电力大学,2017.
[14]田立坤,刘晓宏,李洁.BP神经网络用于大视场显示设备的畸变校正[J].电光与控制,2012,19(12):43-47.
[2]Tsai R Y.A versatile camera calibration technique for high accuracy SD machine vision metrology using off-the-shelf TVcameras and lenses[J].IEEE Jourmal of Robotics Automation.1987,3(4):323-344.
[3]Zhang Z.A flexible new technique for camera calibration[J].IEEE Transactions on Pattern Analysis and Machine Intelligence.2000,22(11):1330-1334.
[4]叶峰,王敏,陈剑东,等.共面点的摄像机非线性畸变校正[J].光学精密工程,2015,23(10):2962-2970.
[5]杨麒,李天伟,黄谦,等.基于直线投影特征的镜头畸变校正方法[J].计算机测量与控制,2018,26(9):264-268.
[6]Song G Y,Lee J W.Correction of radial distortion based on line-fitting[J].International Journal of Control,Automation and Systems,IEEE,2010,18(3):615-621.
[7]白福忠,王建新,杨慧珍,等.视觉测量技术基础[M].北京:电子工业出版社,2013.
[8]Weng J,Cohen P,Herniou M.Camera Calibration with Distortion Models and Accuracy Evaluation,IEEE Trans.Pattern A-nalysis and Machine Intelligence,1992,14(10):965-980.
[9]王红平,曹国华,李振辉,等.基于BP网络的图像畸变校正技术研究[J].长春理工大学学报(自然科学版),2009,32(4):653-656.
[10]江祥奎,范永青,王婉.基于粒子群遗传算法的BP神经网络摄像机标定[J].计算机科学与探索,2014,8(10):1254-1262.
[11]任贝,韩飞,吴坚.基于遗传算法的摄像机标定[J].吉林大学学报(信息科学版),2013,31(4):432-436.
[12]韩啸.基于遗传算法的摄像机内参数标定研究[D].长春:吉林大学,2008.
[13]徐正宏.基于遗传-模拟退火混合算法的泰兴市高压配电网优化规范研究[D].北京:华北电力大学,2017.
[14]田立坤,刘晓宏,李洁.BP神经网络用于大视场显示设备的畸变校正[J].电光与控制,2012,19(12):43-47.