Xianwei Zhu

Robust camera pose estimation from unknown or known line correspondences

We address the model-to-image registration problem with line features in the following two ways. (a) We present a robust solution to simultaneously recover the camera pose and the three-dimensional-to-two-dimensional line correspondences. With weak pose priors, our approach progressively verifies the pose guesses with a Kalman filter by using a subset of recursively found match hypotheses. Experiments show our method is robust to occlusions and clutter. (b) We propose a new line feature based pose estimation algorithm, which iteratively optimizes the objective function in the object space. Experiments show that the algorithm has strong robustness to noise and outliers and that it can attain very accurate results efficiently.

The effects of temperature variation on videometric measurement and a compensation method

When a videometric system operates over a long period, temperature variations in the camera and its environment will affect the measurement results, which cannot be ignored. How to eliminate or compensate for the effects of such variations in temperature is an emergent problem. Starting with the image drift phenomenon, this paper presents an image-drift model that analyzes the relationship between variations in the camera parameters and drift in the coordinates of the image. A simplified model is then introduced by analyzing the coupling relationships among the variations in the camera parameters. Furthermore, a model of the relationship between the camera parameters and temperature variations is established with the system identification method. Finally, several compensation experiments on image drift are carried out, using the parameter-temperature relationship model calibrated with one arbitrary data set to compensate the others. The analyses and experiments demonstrate the feasibility and efficiency of the proposed method. This paper proposed a compensation method for eliminating the effects of temperature variation in the long duration application of image vision.A model of the relationship between the camera parameters and temperature variations is established with the system identification method.Experiments are carried on. The analyses and experiments demonstrate the feasibility and efficiency of the proposed method.

A weighted least squares image matching based target tracking algorithm

Moving target tracking is a basic task in the processing of high speed photography. Despite its widely applications, Correlation tracking method can not adapt to the rotation and zoom of target and results in accumulation of tracking error. The Least Squares Image Matching(LSIM) method which is used in photogrammetry is introduced to moving target tracking, and a Weighted Least Squares Image Matching(WLSIM) based tracking algorithm is proposed. The WLSIM based algorithm sets weights according to the targets shape for the Least-Squares Image Matching Algorithm, as a result matching error produced by the background in the tracking window can be avoided. Experimental results are shown to demonstrate the robustness, efficiency and accuracy of the proposed algorithm.

Salient object detection approach in UAV video

The automatic detection of visually salient information from abundant video imagery is crucial, as it plays an important role in surveillance and reconnaissance tasks for Unmanned Aerial Vehicle (UAV). A real-time approach for the detection of salient objects on road, e.g. stationary and moving vehicle or people, is proposed, which is based on region segmentation and saliency detection within related domains. Generally, the traditional method specifically depends upon additional scene information and auxiliary thermal or IR sensing for secondary confirmation. However, this proposed approach can detect the interesting objects directly from video imagery captured by optical camera fixed on the small level UAV platform. To validate this proposed salient object detection approach, the 25 Hz video data from our low speed small UAV are tested. The results have demonstrated the proposed approach performs excellently in isolated rural environments.

Full-parameter vision navigation based on scene matching for aircrafts

Vision navigation based on scene matching between real-time images and a reference image has many advantages over the commonly used inertial navigation system (INS), such as no cumulative measurement errors for long-endurance flight. Recent developments in vision navigation are mainly used for partial navigation parameters measurements, such as the position and the relative velocity, which cannot meet the requirements of completely autonomous navigation. We present the concept, principle and method of full-parameter vision navigation (FPVN) based on scene matching. The proposed method can obtain the three-dimensional (3D) position and attitude angles of an aircraft by the scene matching for multiple feature points instead of a single point in existing vision navigations. Thus, FPVN can achieve the geodetic position coordinates and attitude angles of the aircraft and then the velocity vector, attitude angular velocity and acceleration can be derived by the time differentials, which provide a full set of navigation parameters for aircrafts and unmanned aerial vehicles (UAVs). The method can be combined with the INS and the cumulative errors of the INS can be corrected using the measurements of FPVN rather than satellite navigation systems. The approach provides a completely autonomous and accurate navigation method for long-endurance flight without the help of satellites.

An Innovative Multi-headed Camera Network: A Displacement-Relay Videometrics Method in Unstable Areas

The subsidence of ground beds occurs frequently in large construction projects such as bridges, dams, high buildings and railways, which may cause damage and sometimes create disasters. Especially along with the establishment and the development of the high-speed railway, tiny subsidence and deformation of the railway beds will result in the decline in the service quality and safety. Therefore, automatic and long-duration subsidence surveillance is becoming increasingly important.

Real-time video exploitation system for small UAVs

The high portability of small Unmanned Aircraft Vehicles (UAVs) makes them play an important role in surveillance and reconnaissance tasks, so the military and civilian desires for UAVs are constantly growing. Recently, we have developed a real-time video exploitation system for our small UAV which is mainly used in forest patrol tasks. Our system consists of six key models, including image contrast enhancement, video stabilization, mosaicing, salient target indication, moving target indication, and display of the footprint and flight path on map. Extensive testing on the system has been implemented and the result shows our system performed well.

Vision navigation for UAV based on scene matching

As a rising navigation technology, vision navigation has many advantages, such as passive measurement, antiinterference, no accumulation of error and comprehensive parameters, etc. It shows a promising application prospects in autonomous navigation for UAV. Based on an efficient, reliable and accurate scene matching, a vision altimeter and 3-D position estimation method are proposed. By matching multiple points between aerial image and reference image, it estimates UAVs position and height according to photogrammetry. To measure UAVs velocity, a mapless speed measurement method which tracks ground features between image sequences is introduced. Flight tests had shown the effectiveness and accuracy of our methods.

Large bulk-yard 3D measurement based on videogrammetry and projected contour aiding

Fast and accurate 3D measurement of large stack-yard is important job in bulk load-and-unload and logistics management. Stack-yard holds its special characteristics as: complex and irregular shape, single surface texture and low material reflectivity, thus its 3D measurement is quite difficult to be realized by traditional non-contacting methods, such as LiDAR(LIght Detecting And Ranging) and photogrammetry. Light-section is good at the measurement of small bulk-flow but not suitable for large-scale bulk-yard yet. In the paper, an improved method based on stereo cameras and laser-line projector is proposed. The due theoretical model is composed from such three key points: corresponding point of contour edge matching in stereo imagery based on gradient and epipolar-line constraint, 3D point-set calculating for stereo imagery projected-contour edge with least square adjustment and forward intersection, then the projected 3D-contour reconstructed by RANSAC(RANdom SAmpling Consensus) and contour spatial features from 3D point-set of single contour edge. In this way, stack-yard surface can be scanned easily by the laser-line projector, and certain regions 3D shape can be reconstructed automatically by stereo cameras on an observing position. Experiment proved the proposed method is effective for bulk-yard 3D measurement in fast, automatic, reliable and accurate way.

Multiple facula targets recognition based on edge and region search

An algorithm of multiple facula targets recognition based on edge and region search in full field of a frame of image is presented. Firstly, the image is segmented by binarization and the burr around the targets is removed by morphology processing. Then every facula targets edge and region is found and numbered in turn. Experimental results on simulated images and real images are shown to validate the presented algorithm.