基于Java+OpenCV技术对几何图像颜色与形状识别
|
摘要
通过Java调用OpenCV视觉库实现几何图像颜色、形状识别。将图像转化为HSV颜色空间,按各颜色H、S、V的取值范围分割图像,对分割图像进行滤波处理、灰度处理、二值化、轮廓提取、去除干扰轮廓、轮廓形状拟合、获取轮廓凸点个数、用凸点距离判断形状,从而实现对各种颜色的三角形、圆、矩形、正方形、多边形等几何图像颜色、形状的识别。该方法可应用于现实生活中,例如识别交通信号灯和信号转向灯等有颜色和形状识别的领域。
This paper uses Java and OpenCV visual library to realize geometric image color and shape recognition. The image is transformed into HSV color space, and the image is segmented according to the range of H, S and V values of each color. The segmented image is processed by filtering, grayscale processing, binarization, contour extraction, removing interference profile, contour shape fitting, obtaining the number of convex points, judging the shape by the distance of convex point, so as to realize the image color and shape recognition. It can be used in real life, such as the recognition of traffic lights and signal turning lights.
This paper uses Java and OpenCV visual library to realize geometric image color and shape recognition. The image is transformed into HSV color space, and the image is segmented according to the range of H, S and V values of each color. The segmented image is processed by filtering, grayscale processing, binarization, contour extraction, removing interference profile, contour shape fitting, obtaining the number of convex points, judging the shape by the distance of convex point, so as to realize the image color and shape recognition. It can be used in real life, such as the recognition of traffic lights and signal turning lights.
引文
[1]周倩,王军,王亮亮.基于OpenCV轮廓逼近的Android简单图形识别[J].福建电脑,2016,32(2):122-123.
[2]于仕琪,刘瑞祯.学习OpenCV[M].北京:清华大学出版社,2009.
[3]阴法名.基于OpenCV图像处理[J].科技信息,2009(32):220.
[4]甘玲,林小晶.基于连通域提取的车牌字符分割算法[J].计算机应用与软件,2011,28(4):336-339.
[2]于仕琪,刘瑞祯.学习OpenCV[M].北京:清华大学出版社,2009.
[3]阴法名.基于OpenCV图像处理[J].科技信息,2009(32):220.
[4]甘玲,林小晶.基于连通域提取的车牌字符分割算法[J].计算机应用与软件,2011,28(4):336-339.