基于最大熵生长检测器的模糊红外图像分割算法
|
摘要
针对图像背景复杂、目标边缘模糊的问题,提出了基于最大熵生长检测器的图像分割算法。根据图像的特点,首先利用多源区域生长法,定位目标,去除背景干扰及噪声;然后通过最大熵原理初步划分目标及背景区域,生成对应区域的图像;分别对这两类区域的图像进行特征训练,得到包含目标区域与背景区域信息的检测器;使用目标与背景的检测器对最大熵分割后的图像进行目标与背景的细分割,最终生成目标图像。实验结果表明,本算法具有良好的目标提取效果。
In this paper,an image segmentation algorithm based on maximum entropy growth detector is proposed to solve the problem of image background complexity and target edge blur. Firstly,it uses the multi-source region growing method to locate the target according to the characteristics of the processed image,and removes the background interference and noise. The maximum entropy principle is utilized to roughly divide the target area and the background,and the image of the target and the background area is generated. Then we train the features of these two types of images and get the detector that contains the information of target area and background area. Using the detector of the target and the background,the image is segmented by the target and the background,and finally the target image is generated. Experimental results show that this algorithm has good effect on object extraction.
In this paper,an image segmentation algorithm based on maximum entropy growth detector is proposed to solve the problem of image background complexity and target edge blur. Firstly,it uses the multi-source region growing method to locate the target according to the characteristics of the processed image,and removes the background interference and noise. The maximum entropy principle is utilized to roughly divide the target area and the background,and the image of the target and the background area is generated. Then we train the features of these two types of images and get the detector that contains the information of target area and background area. Using the detector of the target and the background,the image is segmented by the target and the background,and finally the target image is generated. Experimental results show that this algorithm has good effect on object extraction.
引文
[1] VEENMAN C,REINDERS M,BACKER E. A cellular coevolutionary algorithm for image segmentation[J]. IEEE Transaction on Image Processing,2003,l2(3):304-316
[2]王艳娟,陈晓红,邹丽.图像感兴趣区域自动提取算法[J].科学技术与工程. 2007,(12):2867-2871.
[3]陈玉璞.基于神经网络模型的彩色图像分割新方法[D].北京:北京工业大学,2013.
[4]邹小林,冯国灿.融合视觉模型和最大熵的阈值分割算法[J].科学技术与工程. 2013,(6):1496-1501.
[5]孔英会,陈咨彤,车辚辚.基于关键子区域及特征提取的表情识别[J].科学技术与工程. 2017,(34):257-262.
[6]冯威,高天德,闫永胜.基于灰度不均匀图像边缘检测算法改进[J].科学技术与工程. 2010,(36):9010-9013.
[7]朱占龙,唐志远,李明亮,等.二维Arimoto熵阈值分割法中参数的智能选择方法[J].科学技术与工程. 2017,(19):191-196.
[8] QIYAO YU,DAVID CLAUSI. IRGS:Image Segmentation Using Edge Penalties and Region Growing[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2008,30(12):2126-2139
[9] ZHAO Y Q,WANG X H,WANG X F,et al. Retinal vessels segmentation based on level set and region growing[J].Pattern Recognition,2014,47(7):2437-2446
[10] ASGARI T S,LIU Y,MILES B,et al. Geometry Based Pectoral Muscle Segmentation from MLO Mammogram Views.[J]. IEEE Transactions on Biomedical Engineering,2017,PP(99):1-1
[11]马腾,杨大利,侯凌燕.基于双边滤波与区域生长的织物疵点检测[J].北京信息科技大学学报:自然科学版. 2017,(1):82-85.
[12]李迎春,付兴建,薛琴.基于RGB熵的非结构化道路分割方法[J].计算机工程与设计,2017,38(6):1570-1574.
[13]邢丁亮,郝占庆.最大熵原理及其在生态学研究中的应用[J].生物多样性. 2011,(3):295-302.
[14]任群.基于信息熵的图像处理技术[J].大庆师范学院学报. 2017,(3):26-29.
[15]朱桂英,张瑞林.信息熵在图像处理中的应用[J].丝绸. 2006,(12):34-36.
[16]陶思然,陈性义,张海涛.优化区域生长的高分辨率影像道路精确提取[J].科学技术与工程. 2017,(33):136-143.
[17]陈琳,唐忠,崔昊杨.电气设备红外测温技术的实现[J].电测与仪表. 2013,(4):64-68.
[18]田琳,苗加庆.图像分割在等离子体加料实验中的应用研究[J].激光杂志,2014(12):50-53.
[2]王艳娟,陈晓红,邹丽.图像感兴趣区域自动提取算法[J].科学技术与工程. 2007,(12):2867-2871.
[3]陈玉璞.基于神经网络模型的彩色图像分割新方法[D].北京:北京工业大学,2013.
[4]邹小林,冯国灿.融合视觉模型和最大熵的阈值分割算法[J].科学技术与工程. 2013,(6):1496-1501.
[5]孔英会,陈咨彤,车辚辚.基于关键子区域及特征提取的表情识别[J].科学技术与工程. 2017,(34):257-262.
[6]冯威,高天德,闫永胜.基于灰度不均匀图像边缘检测算法改进[J].科学技术与工程. 2010,(36):9010-9013.
[7]朱占龙,唐志远,李明亮,等.二维Arimoto熵阈值分割法中参数的智能选择方法[J].科学技术与工程. 2017,(19):191-196.
[8] QIYAO YU,DAVID CLAUSI. IRGS:Image Segmentation Using Edge Penalties and Region Growing[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2008,30(12):2126-2139
[9] ZHAO Y Q,WANG X H,WANG X F,et al. Retinal vessels segmentation based on level set and region growing[J].Pattern Recognition,2014,47(7):2437-2446
[10] ASGARI T S,LIU Y,MILES B,et al. Geometry Based Pectoral Muscle Segmentation from MLO Mammogram Views.[J]. IEEE Transactions on Biomedical Engineering,2017,PP(99):1-1
[11]马腾,杨大利,侯凌燕.基于双边滤波与区域生长的织物疵点检测[J].北京信息科技大学学报:自然科学版. 2017,(1):82-85.
[12]李迎春,付兴建,薛琴.基于RGB熵的非结构化道路分割方法[J].计算机工程与设计,2017,38(6):1570-1574.
[13]邢丁亮,郝占庆.最大熵原理及其在生态学研究中的应用[J].生物多样性. 2011,(3):295-302.
[14]任群.基于信息熵的图像处理技术[J].大庆师范学院学报. 2017,(3):26-29.
[15]朱桂英,张瑞林.信息熵在图像处理中的应用[J].丝绸. 2006,(12):34-36.
[16]陶思然,陈性义,张海涛.优化区域生长的高分辨率影像道路精确提取[J].科学技术与工程. 2017,(33):136-143.
[17]陈琳,唐忠,崔昊杨.电气设备红外测温技术的实现[J].电测与仪表. 2013,(4):64-68.
[18]田琳,苗加庆.图像分割在等离子体加料实验中的应用研究[J].激光杂志,2014(12):50-53.