3-D object reconstruction using stereo and motion

2010 IEEE Intelligent Vehicles Symposium, 2010

This paper presents a method for improving the disparity values obtained with a stereo camera by applying an iconic Kalman filter and known ego-motion. The improvements are demonstrated in an application of determining the free space in a scene as viewed by a vehicle-mounted camera. Using disparity maps from a stereo camera and known camera motion, the disparity maps are first filtered by an iconic Kalman filter, operating on each pixel individually, thereby reducing variance and increasing the density of the filtered disparity map. Then, a stochastic occupancy grid is calculated from the filtered disparity map, providing a top-down view of the scene where the uncertainty of disparity measurements are taken into account. These occupancy grids are segmented to indicate a maximum depth free of obstacles, enabling the marking of free space in the accompanying intensity image. Even without motion of the camera, the quality of the disparity map is increased significantly. Applications of the intermediate results are discussed, enabling features such as motion detection and quantifying the certainty of the measurements. The evaluation shows significant improvement in disparity variance and disparity map density, and consequently an improvement in the application of marking free space.

Elektronika Ir Elektrotechnika, 2015

Machine Vision - Applications and Systems, 2012

Remote Sensing

Binocular disparity and motion parallax are the most important cues for depth estimation in human and computer vision. Here, we present an experimental study to evaluate the accuracy of these two cues in depth estimation to stationary objects in a static environment. Depth estimation via binocular disparity is most commonly implemented using stereo vision, which uses images from two or more cameras to triangulate and estimate distances. We use a commercial stereo camera mounted on a wheeled robot to create a depth map of the environment. The sequence of images obtained by one of these two cameras as well as the camera motion parameters serve as the input to our motion parallax-based depth estimation algorithm. The measured camera motion parameters include translational and angular velocities. Reference distance to the tracked features is provided by a LiDAR. Overall, our results show that at short distances stereo vision is more accurate, but at large distances the combination of pa...

Electronics Letters, 2008

An algorithm for refinement of disparity (depth) map from stereoscopic sequences is proposed. The method is based on estimation of ego-motion parameters of a camera system and 3D scene frame-by-frame prediction.

2016 IEEE International Conference on Automatic Control and Intelligent Systems (I2CACIS), 2016

In this paper, a 3D scene reconstruction by using stereo vision is presented. Stereo camera parameters from the camera calibration process and disparity map are two important parameters to obtain an accurate result for the 3D reconstruction. 3D reconstruction process generates the coordinates of world points (point cloud). From the information of the point cloud, the interested object is segmented out. Additionally, the noises left in the image is eliminated. The experimental result obtained shows that only the background and the interested object from the overall scene appeared in the processed image. The center of gravity is also determined as the reference value for the robot navigation. An estimation error model is also introduced in this study to increase the accuracy of the distance measurement by using the developed stereo vision system from the mobile robot. This experiment provides a foundation for navigation and object tracking for a mobile robot.

1999

In this paper, we examine and present the development of a depth map acquisition system with real-time characteristics based on a binocular active vision system. In order to obtain depth information we present a method that combines stereo matching with mechanical activity, reducing the time spent to perform the correspondence between the image points on the left and right images. Controlling the cameras' vergence or the baseline distance it is possible to change continuously the xation point in the space and, at the same time, to select points with correspondent image projections. Computing the distance from those points to the vision system, is possible to obtain a dense relative map of the scene. The correspondence is established based on similarity measures between image regions. This measures are performed by operators with characteristics that makes this method suitable for parallel implementation. Since the depth information is relative, the calibration of the active vision system is minimal.

IEEE Transactions on Systems, Man, and Cybernetics, 1989

The development of a practical binocular stereo approach for the purpose of extracting depth information is considered. Accurate measurement of depth is required for many robotic tasks such as object recognition, manipulation, assembly, and obstacle avoidance. The ability to produce precise depth measurements over a wide range of distances and the passivity of the approach requiring simple and commonly available sensors, make binocular stereo an attractive approach. Yet rigid input requirements and high computational costs have prevented stereo approaches from being adopted for practical robotic applications. For effective use in this application domain, a stereo system must seek a balance between input flexibility, output sufficiency, and computational complexity. At the heart of a binocular stereo approach lies the task of stereo matching. The importance of semantic content and stability of the matching primitive is emphasized and the use of homogeneous regions as features in stereo matching is introduced. The region-based matching is more accurate than those using edge-based primitives since regions have higher discrimination capability. Also, region-based approaches are more efficient since there are fewer features to be matched. However, while the use of regions makes stereo matching more accurate, reliable, and efficient than edge-based matching, region-based matching processes typically yield coarse disparity maps. The generation of accurate and finer resolution disparity maps (and subsequently depth measurements) can be better accomplished using edge-based techniques. Both regions and edges play important, but somewhat complementary, roles in the binocular stereo process. It is, therefore, critical that an efficient and robust stereo system utilize the most appropriate set of primitives at each stage of the process. A hierarchical stereo approach that exploits and integrates the power of different primitives at appropriate stages of the stereo process is proposed. Several experiments to evaluate the performance of a region-based stereo matcher and a disparity and depth generation module are presented. It is shown that the approach is efficient, robust, and successful in generating depth measurements within f 5 percent accuracy.

Proceedings of gruppo telecomunicazioni e tecnologie dell’informazione (GTTI) meeting, 2010

3D video applications require the acquisition of high quality depth information in real time. This issue can be solved using stereo vision systems or time-of-flight (ToF) cameras. Both solutions present critical issues, that can be overcome by their combined use. A heterogeneous acquisition system is considered in this paper, made of two high resolution standard cameras (stereo pair) and one ToF camera. The ToF range camera technology is quite new, therefore a brief description of how a ToF range cameras works is firstly given, ...