CAJ | 학술논문

为了实现运动目标的检测,提出使用一种基于自适应阈值和主成分分析(Principal Components Analysis,PCA)的运动目标检测方法.首先,拍摄一组静态环境下无运动物体的图像以得到PCA使用的变换矩阵.通过变换矩阵,连续的图像可以被映射到变换空间;相反,变换后的图像通过逆变换也可以被还原.通过计算原图像和还原图像之间的欧氏距离,可以实现运动目标的检测.欧式距离大于阈值的图像区域被认为属于运动目标.算法通过阈值的动态调整得到自适应阈值,从而在很大程度上实现对光照等环境条件变化的补偿.实验结果表明,此方法具有较好的鲁棒性和有效性.
위료실현운동목표적검측,제출사용일충기우자괄응역치화주성분분석(Principal Components Analysis,PCA)적운동목표검측방법.수선,박섭일조정태배경하무운동물체적도상이득도PCA사용적변환구진.통과변환구진,련속적도상가이피영사도변환공간;상반,변환후적도상통과역변환야가이피환원.통과계산원도상화환원도상지간적구씨거리,가이실현운동목표적검측.구식거리대우역치적도상구역피인위속우운동목표.산법통과역치적동태조정득도자괄응역치,종이재흔대정도상실현대광조등배경조건변화적보상.실험결과표명,차방법구유교호적로봉성화유효성.
In order to detect motion object, a moving object detection method based on adaptive threshold and Principal Components Analysis (PCA) is presented. First, a set of images of the static environment without motion object is captured to obtain the transformation matrix that used in PCA. By means of this matrix, the successive images are projected in the transformation space. On the contrary, the transformed image can also be recovered using inverse transformation. By evaluating the Euclidean distance between the original and recovered images, the motion detection is performed. The image regions whose Euclidean distance is greater than a threshold is considered like belonging to motion objects. By dynamically adjusting threshold, the algorithm can obtain the adaptive threshold that compensates to a great extent, illumination and other environmental conditions variations. The experimental results show that, the method has better robustness and effectiveness.