基于OpenCV的复杂环境下图像二值化方法
|
摘要
机械式电表读数图像易受光照不均、反光等因素影响,导致传统二值化方法效果不佳。针对以上问题,基于OpenCV开源视觉库,提出一种结合改进的Canny边缘检测算法、全局阈值与局部自适应阈值的图像二值化方法。从图像自身特点出发选取Canny算子最佳全局双阈值,并进行局部动态阈值修正实现准确的边缘检测。综合图像边缘信息、全局阈值与局部自适应阈值实现图像二值化。实验表明在光照不均、反光等因素影响下,该方法取得了良好的二值化效果。
The mechanical meter numeral images often acquired under non-uniform illumination and reflection.The traditional methods can't binarize them correctly.To solve the problem,a binarization method based on OpenCV by combining Canny's edge detection with global threshold and local adaptive threshold is proposed.In order to improve the edge detection effect,the high and low threshold parameters for the Canny operator is determined according to the characteristics of the meter numeral images,and the local dynamic threshold is involved for correction.The binarization method is achieved by the fusion of the image edge,global threshold and local adaptive threshold.The test shows that the method can achieve good effect under non-uniform illumination and reflection.
The mechanical meter numeral images often acquired under non-uniform illumination and reflection.The traditional methods can't binarize them correctly.To solve the problem,a binarization method based on OpenCV by combining Canny's edge detection with global threshold and local adaptive threshold is proposed.In order to improve the edge detection effect,the high and low threshold parameters for the Canny operator is determined according to the characteristics of the meter numeral images,and the local dynamic threshold is involved for correction.The binarization method is achieved by the fusion of the image edge,global threshold and local adaptive threshold.The test shows that the method can achieve good effect under non-uniform illumination and reflection.
引文
[1]张文杰,熊庆宇,张家齐,等.基于视觉显著性的指针式仪表读数识别算法[J].计算机辅助设计与图形学学报,2015,27(12):2282-2295.
[2]郭佳,刘晓玉,吴冰,等.一种光照不均匀图像的二值化方法[J].计算机应用与软件,2014,31(3):183-186,202.
[3]吴佳鹏,杨兆选,韩东,等.基于小波和Otsu法的二维条码图像二值化[J].计算机工程,2010,36(10):190-192.
[4]陈强,朱立新,夏德深.结合Canny算子的图像二值化[J].计算机辅助设计与图形学学报,2005(6):1302-1306.
[5]李鹏,蔡玉雷,陈金立.采用椭圆算法和Canny算子的超声TOF估计方法[J].仪器仪表学报,2014,35(10):2385-2391.
[6]计春雷,冯伟,黎明,等.一种动态阈值加填补的指纹图像二值化算法[J].计算机仿真,2011,28(7):258-261,289.
[7]张帆,彭中伟,蒙水金.基于自适应阈值的改进Canny边缘检测方法[J].计算机应用,2012,32(8):2296-2298.
[8]朱海洋,徐根玖,李元晨,等.基于图像分块的局部区域动态阈值选取方法[J].计算机与现代化,2016(11):53-57.
[9]谭凯,程效军.双阈值法地面激光点云强度图像边缘提取[J].同济大学学报(自然科学版),2015,43(9):1425-1431.
[10]董昱,高云波,刘翔.改进的遗传算法在Canny算子阈值选取中的应用[J].兰州交通大学学报,2014,33(6):1-5.
[11]赵岩,周百灵,陈贺新.一种改进的基于Canny算子边缘检测算法[J].吉林大学学报(理学版),2012,50(4):740-744.
[12]孙智鹏,邵仙鹤,王翥,等.改进的自适应Canny边缘检测算法[J].电测与仪表,2016,53(6):17-21.
[13]王佳,李波,徐其志.边缘检测中局部区域的动态阈值选取方法[J].计算机应用研究,2010,27(2):772-774.
[14]吴丽丽,余春艳.基于Sobel算子和Radon变换的车牌倾斜校正方法[J].计算机应用,2013,33(S1):220-222.
[15]沈德海,侯建,鄂旭.基于改进的Sobel算子边缘检测算法[J].计算机技术与发展,2013,23(11):22-25.
[16]童立靖,张艳,舒巍,等.几种文本图像二值化方法的对比分析[J].北方工业大学学报,2011,23(1):25-33.
[17]李康顺,阮丹丹,张文生.基于演化算法的图像二值化算法研究[J].计算机应用研究,2012,29(1):348-351.
[2]郭佳,刘晓玉,吴冰,等.一种光照不均匀图像的二值化方法[J].计算机应用与软件,2014,31(3):183-186,202.
[3]吴佳鹏,杨兆选,韩东,等.基于小波和Otsu法的二维条码图像二值化[J].计算机工程,2010,36(10):190-192.
[4]陈强,朱立新,夏德深.结合Canny算子的图像二值化[J].计算机辅助设计与图形学学报,2005(6):1302-1306.
[5]李鹏,蔡玉雷,陈金立.采用椭圆算法和Canny算子的超声TOF估计方法[J].仪器仪表学报,2014,35(10):2385-2391.
[6]计春雷,冯伟,黎明,等.一种动态阈值加填补的指纹图像二值化算法[J].计算机仿真,2011,28(7):258-261,289.
[7]张帆,彭中伟,蒙水金.基于自适应阈值的改进Canny边缘检测方法[J].计算机应用,2012,32(8):2296-2298.
[8]朱海洋,徐根玖,李元晨,等.基于图像分块的局部区域动态阈值选取方法[J].计算机与现代化,2016(11):53-57.
[9]谭凯,程效军.双阈值法地面激光点云强度图像边缘提取[J].同济大学学报(自然科学版),2015,43(9):1425-1431.
[10]董昱,高云波,刘翔.改进的遗传算法在Canny算子阈值选取中的应用[J].兰州交通大学学报,2014,33(6):1-5.
[11]赵岩,周百灵,陈贺新.一种改进的基于Canny算子边缘检测算法[J].吉林大学学报(理学版),2012,50(4):740-744.
[12]孙智鹏,邵仙鹤,王翥,等.改进的自适应Canny边缘检测算法[J].电测与仪表,2016,53(6):17-21.
[13]王佳,李波,徐其志.边缘检测中局部区域的动态阈值选取方法[J].计算机应用研究,2010,27(2):772-774.
[14]吴丽丽,余春艳.基于Sobel算子和Radon变换的车牌倾斜校正方法[J].计算机应用,2013,33(S1):220-222.
[15]沈德海,侯建,鄂旭.基于改进的Sobel算子边缘检测算法[J].计算机技术与发展,2013,23(11):22-25.
[16]童立靖,张艳,舒巍,等.几种文本图像二值化方法的对比分析[J].北方工业大学学报,2011,23(1):25-33.
[17]李康顺,阮丹丹,张文生.基于演化算法的图像二值化算法研究[J].计算机应用研究,2012,29(1):348-351.