摘要
圆盘靶标在摄像机标定及无人机侦察任务中目标的定位方面有着广泛的应用。针对目前图像角点检测的局限性,提出了一种自动提取红蓝彩色圆盘靶目标中心角点的检测方法。该方法利用HSV(Hue,saturation,value)颜色空间对红蓝彩色圆盘进行颜色分离,获得具有明显对称性和对比度强烈的灰度图像。然后运用提出的新型对称算子计算每个像素点的响应值作为角点的判断依据,对求得的角点在其邻域内抑制红蓝像素个数差来确保角点的对称性。最后通过Matlab软件上的仿真和现场可编程门阵列(FPGA)硬件上的实现,证明了该方法易于硬件化实现,实时性强,抗噪性能好,对图像旋转,伸缩和尺度变换都有良好的稳健性。
Target disc images are widely used in camera calibration and target location for unmanned （UAV） scout task, so a method of automatically extracting feature point coordinates is proposed for the color on the two- dimensional （2D） center of the disc to solve the problems of present corner detection algorithms. The operation of color separation for the red and blue discs is completed with the method of hue-saturation-value （HSV） volor space in order to get a gray image which has obvious symmetry and strong contrast. According to the corner points of the candidate red and blue pixels in the neighborhood with the distance less than a threshold number, the exact value of the corner is determined, and the response value of extracted large corner point is applied as a candidate by using a new type of symmetric operator. Experiment results of Matlah simulation and hardware implementation based on the field programmable gate array （FPGA） platform show that the proposed method is easy to implement in hardware technology, real-time and has good anti-noise ability. Also it has a good robustness to image rotation, stretching and scale transformation.
出处
《激光与光电子学进展》
CSCD
北大核心
2014年第3期82-91,共10页
Laser & Optoelectronics Progress
基金
国家自然科学基金(61101119)
江苏省普通高校研究生科研创新基金(CXZZ12_0183)
江苏省自然科学基金(BK2011699)
