CAJ | 학술논문

针对机载光电系统目标检测的特点,提出了一种基于平台运动补偿与高斯混合模型相结合的下视场景运动目标检测方法。通过图像配准实现平台的运动补偿,然后建立自适应高斯混合模型,经背景差法提取运动目标,采用多次假设检验的方式剔除提取出来的潜在目标中的干扰目标。实验结果证明,该方法对机载运动平台下运动目标检测快速有效。
침대궤재광전계통목표검측적특점,제출료일충기우평태운동보상여고사혼합모형상결합적하시장경운동목표검측방법。통과도상배준실현평태적운동보상,연후건립자괄응고사혼합모형,경배경차법제취운동목표,채용다차가설검험적방식척제제취출래적잠재목표중적간우목표。실험결과증명,해방법대궤재운동평태하운동목표검측쾌속유효。
Based on the characteristics of the objects detection in airborne optoelectronic systems,an algorithm for detecting moving object in downward-looking systems,which combines the platform motion compensation and Gaussian mixture models（GMM）,is proposed in this paper.Firstly,this algorithm compensates the motion for the airborne platform by image registration,then constructs adaptive Gaussian mixture models,and extracts the moving objects using background-subtracting method.At last,the interfering objects are eliminated by multilevel hypothesis testing.Experimental results demonstrate that the proposed algorithm is fast and effective for the objects detection under the airborne platform.