《动态场景中的目标检测算法的研究》由会员分享，可在线阅读，更多相关《动态场景中的目标检测算法的研究（63页珍藏版）》请在金锄头文库上搜索。
1、合肥工业大学硕士学位论文动态场景中的目标检测算法的研究姓名：吴晖申请学位级别：硕士专业：信号与信息处理指导教师：蒋建国2011-03动态场景中的目标检测算法的研究动态场景中的目标检测算法的研究 摘摘 要要 运动目标检测是计算机视觉中的基础性课题，在人类社会的多个领域内具有广泛的应用前景和研究价值。目标检测可以分为静态场景和动态场景两种情况。其中，动态场景中的运动目标检测除了具有静态场景中的运动目标检测所要解决的问题，还要去除摄像机自运动的影响，因此它的研究难度更大，目前发展仍不成熟。本文特别针对动态场景的目标检测问题，将动态场景划分为二维和三维两种类型，并研究了各自情况下的解决算法。 二维场景
2、下的运动目标检测二维场景下的运动目标检测。我们引入了一种通用性强的全局运动模型，用特征点匹配的方法计算全局运动参数，并进行运动补偿，从而检测出运动目标。该算法经过对多组实拍和标准序列的测试，证明能够实时精确地检测出二维场景中的运动目标。该部分主要新颖之处有： （1）分析了不同摄像机运动情况下和不同场景类型下的图像全局运动，并且引入八参数运动模型来描述图像全局运动。 （2）针对 SIFT 特征点匹配速度慢的缺陷，实现了基于特征点预测的检测和匹配策略，在不降低 SIFT 本身性能的前提下，极大的提高了算法效率。 （3）基于残差图像的特征点更新策略，即时更新特征点，在提高算法效率的同时，保证了算法鲁
3、棒性。 三维场景下的运动目标检测三维场景下的运动目标检测。该算法使用多视角几何学的知识进行场景点的三维位置重构，从而实现运动目标检测。该算法在室内实拍的长序列三维场景视频上进行验证，实验结果证明，该算法对三维场景中的目标检测结果具有良好的鲁棒性和准确性。此部分的新颖之处如下： （1）在运动目标检测中引入多视角几何学的相关理论。结合极几何和三视角几何，为特征点分配动态性概率。 （2）提出三维位置一致性约束，检测出在连续的三帧图像中，三维位置发生变化的动态点。 关键词关键词： 运动目标检测、极几何、三维重构、运动估计和补偿、特征匹配、动态场景Moving Objects Detection in
4、Dynamic Scenes Abstract Moving objects detection has been one of the most basic issues in Computer Vision, and in various social areas, it shows great prospect for application, such as intelligent surveillance system, robot industries, medical diagnosis, self-guided missile and so on. Because of its
5、 profound civil and military values, researching moving object detection received high emphasis in the developing process of Computer Vision in recent years. And the fact that detecting in dynamic scenes includes both problems related to static scenes and problems introduced by camera ego-motion, in
6、creases research difficulty of moving object detection in dynamic scenes, whose development still has a long way to run. Therefore, in this thesis, we divide dynamic scenes into two major categories and research the moving object algorithm in these two cases respectively. Detection in 2D scenes. We
7、introduce a widely applicable global motion model, and solve the motion compensation problem through feature points matching. The algorithm has been tested on several real shot sequences and standard sequences, and results shows that our algorithm is able to accurately achieve motion compensation an
8、d detect motion in real time performance. Our original work can be summarized as: (1) Analyze characteristics of global motion in various types of scenes, and introduce eight-parameter model which is capable to model multiple scenes. (2) As to the shortcoming of SIFT-low matching rate, we propose a
9、feature searching and matching strategy based on position estimation. This can improve the efficiency while not reducing performance of original SIFT. (3) Feature points update strategy based on residual image. This update strategy makes sure that features can be updated frame by frame and thus the
10、running speed increases. Detection in 3D scenes. This algorithm employs multiple views geometry to detect dynamic points in each frame. Two different geometric constraints are used in the process. The algorithm is tested on long-time indoor sequences and results show that it can robustly and correct
11、ly detect moving objects in 3D scenes. Original work of this part can be summarized as: (1) Introduce multiple views geometry into motion detection. Combining epipolar geometry and three-view geometry, the features are assigned with dynamic probabilities. (2) Propose 3D position consistency constrai
12、nt: in 3 consecutive frames, reconstructed points with changing 3D position has a high dynamic probability. Keywords: Moving Object Detection, Epipolar Geometry, 3D Reconstruction, Motion Estimation and Motion Compensation, Feature Matching, Dynamic Scene插图清单插图清单 图 2. 1 摄像机坐标系. 8 图 2. 2 双目几何 . 11 图 2. 3 极平面和极线. 11 图 2. 4 极几何实例 1. 12 图 2. 5 极几何实例 2. 13 图 2. 6 全局运动与局部运动 . 15 图 3. 1 直线的 RANSAC 估计. 28 图 3. 2 算法流程图 . 29 图 3. 3 Tf 值对算法性能的影响 . 32 图 3. 4 标记前景块示意图 . 32 图 3. 5 序列 1 的实验结果 . 34 图 3. 6 序列 2 的实验结果 . 34 图 3. 7 序列 3 的实验结果 . 34 图 3. 8 序列 4 的实验结果 . 35 图 4. 1 运动参数四组解的几何表示 . 40 图 4. 2 三角原理图 .
