Yun Yuan

Superresolution Reconstruction of Video Sequence Using a Coarse-to-Fine Registration and Optimal Interpolation Strategy

Abstract In this paper, a novel video superresolution (SR) reconstruction approach using a coarse-to-fine registration and optimal interpolation strategy is proposed. Firstly, initial correspondence between two consecutive frames is identified by scale invariant feature transform (SIFT) features under the homography transformation model. Then, a least square matching (LSM) method is employed to achieve a high-precision estimation of homography. Finally, optimal interpolation using the iterative back projection-based deblurring method was designed to obtain a high-resolution grid interpolation and to improve the visual quality of the final SR results. The experiment results demonstrate that the proposed method provides an available SR approach in engineering applications.

Projection-aided videometric method for shape measurement of large-scale bulk material stockpile

Large-scale stockpiles are a common means for the storage of bulk material and their shape measurement is a prerequisite for effective transportation and logistics management. This paper proposes a projection-aided videometric method that is based on stereo vision and makes use of projections and epipolar geometry constraints to solve the correspondence problem. A prototype measuring system was developed that can work in real time and achieve measuring accuracy of 0.71 mm in terms of the flatness deviation and 0.54 mm in terms of the depth error, as proven by the experimental results.

Large stack-yard three-dimensional measurement based on videogrammetry and projected-contour scanning

Fast and reliable three-dimensional (3-D) measurement of large stack yards is an important job in bulk load-and-unload operations and logistics management. Traditional noncontacting methods, such as LiDAR and photogrammetry, witness difficulties of complex and irregular shape, single texture and weak reflectivity, and so on. In this paper, we propose a videogrammetry and projected-contour scanning method. The surface of a stack yard can be scanned easily by a laser-line projector, and its 3-D shape can be reconstructed automatically by stereo cameras. There are two main technical contributions of this method: 1. corresponding-point matching in stereo imagery based on image gradient and epipolar line; and 2. single projected-contour extraction under constraint of homography and RANSAC (random sampling consensus). The proposed method has been tested by 3-D-reconstruction experiments of sand tables in indoor and outdoor conditions, which showed that about five contours were reconstructed per second on average, and moving-distance error of a standard slab was less than 0.4 mm in the worst direction of the videogrammetric system. In conclusion, the proposed method is effective for 3-D shape measurement of stack yards in a fast, reliable and accurate way.

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.

Theodolite-camera videometrics system based on total station

A novel measuring system, named Theodolite-camera Videometrics System (TVS) based on total station, has been introduced in this paper, and the concept of theodolite-camera which is the key component of TVS has been proposed, it consists of non-metric camera and rotation platform generally, and can rotate horizontally and vertically. TVS based on total station is free of field control points, and the fields of view of its theodolite-cameras are nonfixed, thus TVS is qualified for targets with wide moving range or big structure. Theodolite-camera model has been analyzed and presented in detail in this paper. The calibration strategy adopted has been demonstrated to be accurate and feasible by both simulated and real data, and TVS has also been proved to be a valid, reliable, precise measuring system, and living up to expectations.