摘要
由于受到物体与摄像机三维空间相对位置关系的影响,摄像机获取的图像存在透视畸变,影响图像特征提取、识别等后续处理,为此提出一种通过分层矫正和Radon变换相结合的矫正技术。利用射影几何知识,将畸变的图像矫正为满足仿射变换的图像,再将仿射变换矫正为相似变换,利用Radon变换矫正得到标准图像。仿真实验结果表明,该算法可行性高、稳定性好,能推广应用于目标识别等领域。
When the images collected from camera have perspective distortion,it can affect the following work such as the feature extraction and object recognition due to the influence of the apace position between the object and the camera.So a algorithm of combining the stratification rectification and Radon transformation is proposed to solve this problem.Firstly,the distortion image is converted to the image,which satisfy affine transformation,then upgrading to similarity transformation by using the projective geometry.At last,Radon transformation is used to obtain the standard image.Experimental results indicate that the algorithm not only is feasibility but also have the high stability and it can be applied to the areas of the target identification.
出处
《计算机工程与设计》
CSCD
北大核心
2010年第13期3108-3110,3131,共4页
Computer Engineering and Design
基金
教育部博士点基金项目(200805321020)
关键词
平面图像
射影畸变
仿射矫正
度量矫正
RADON变换
plane-image
projection distortion
affine rectification
metric rectification
Radon transformation
