摘要
角点提取在图像匹配、相机标定、模式识别和三维重建等方面有着极其重要的应用,是当前图像研究的热点。在阐述经典的Harris算法基础上,分析了其优缺点。针对Harris算法受随机噪声影响较大和运算速度慢的局限,在采用Sobel梯度算子基础上,提出了基于噪声检测的改进算法,实验表明噪声检测算子可有效检出随机噪声,证明了方法的有效性与可行性。
Corner extraction plays an important role in image matching,camera calibration,pattern recognition and three-dimensional reconstruction and is regarded as an hot spot in the research of image processing.Here the advantages and disadvantages of Harris algorithm are analyzed based on the statement of its principle.Aiming at the limits that Harris algorithm is gravely suffered from image noise and slow calculation speed,the new improved algorithm is presented based on noise detection along with Sobel gradient operator.Further,the effectiveness and feasibility of the algorithm are proved and through controlled trials.
出处
《电脑开发与应用》
2010年第8期29-31,共3页
Computer Development & Applications
