Moving photogrammetry is an application of close range photogrammetry in industrial measurement to realize threedimensional coordinate measurement within large-scale volume. This paper describes an approach of relaxation matching algorithm applicable to moving photogrammetry according to the characteristics of accurate matching result of different measuring images. This method uses neighborhood matching support to improve the matching rate after coarse matching based on epipolar geometry constraint and precise matching using three images. It reflects the overall matching effect of all points, that means when a point is matched correctly, the matching results of those points round it must be correct. So for one point considered, the matching results of points round it are calculated to judge whether its result is correct. Analysis indicates that relaxation matching can eliminate the mismatching effectively and acquire 100% rate of correct matching. It will play a very important role in moving photogrammetry to ensure the following implement of ray bundle adjustment.
An optical vortex is a beam of light with phase varying in a corkscrew-like manner along its direction of propagation and so has a helical wavefront. When such a vectorial vortex beam and the Gaussian beam with orthogonal polarization are focused by low NA lens, the Gaussian component causes a focal intensity distribution with a solid center and the vortex component causes a donut distribution with hollow dark center. The shape of the focus can be continuously varied by continuously adjusting the relative weight of the two components. Flat top focusing can be obtained under appropriate conditions. It is demonstrated through experiments with a liquid crystal spatial light modulator in such a beam, that flattop focus can be obtained by vectorial vortex beams with topological charge of +1 to achieve beam shaping vortex.
In this study we analyze one of a CFD for timing discrimination. Walk error, drift and precision are the three performance parameters of timing discrimination. The walk error is the most important error type generally. Firstly, we divided the waveform into two types. One is the Gaussian waveform distribution which has three parameters: amplitude, mean, and the pulse width; and the other is Rayleigh waveform distribution which has two parameters: mean and pulse. We analyzed different situations with their changing parameter, and the drift value of time can be obtained for each parameter changing.
To the question of measuring the moving object pose, a high speed and high synchronization precision spatial pose measurement system based on optical measurement was designed. The system is more convenient and more accurate. In order to realize the measurement method, a Synchronous controller was used to keep the moving object and the pose measurement system based on binocular vision model synchronized. The system can record the course with high synchronization precision. Geometry constraint relation of the special markets and optimization algorithm based on coordinates of multi-points were used in the pose algorithm of the moving object. Experimental results and theoretical analysis prove that the pose measurement method is correct and reliable. The frequency of the pose measurement system is 100 frames per second. The error of the pose angle is less than 0.05°. The pose measurement system satisfies the requirements of pose measuring in ground simulation test.
Positioning error of robot is a main factor of accuracy of flexible coordinate measuring system which consists of universal industrial robot and visual sensor. Present compensation methods for positioning error based on kinematic model of robot have a significant limitation that it isn’t effective in the whole measuring space. A new compensation method for positioning error of robot based on vision measuring technique is presented. One approach is setting global control points in measured field and attaching an orientation camera to vision sensor. Then global control points are measured by orientation camera to calculate the transformation relation from the current position of sensor system to global coordinate system and positioning error of robot is compensated. Another approach is setting control points on vision sensor and two large field cameras behind the sensor. Then the three dimensional coordinates of control points are measured and the pose and position of sensor is calculated real-timely. Experiment result shows the RMS of spatial positioning is 3.422mm by single camera and 0.031mm by dual cameras. Conclusion is arithmetic of single camera method needs to be improved for higher accuracy and accuracy of dual cameras method is applicable.
In this paper, a new coordinate system calibration is proposed in order to define the dependence of position between object body coordinate and camera coordinate which can be used in object measurement by the formation of image. Due to existing the fabrication and installation error, it is difficult to make system parallel between object body coordinate and camera coordinate. To resolve it, the deviation of the two coordinate system is demanded to measure detailed. The deviation compensation in the image processing software can ensure the accuracy the pitch angle and azimuth in the destination image of camera measurement system. In order to definite the position, a base coordinate system of theodolite is set by mutual-space measurement principle. After the measurement of theodolite system, a transformation matrix of the base coordinate system can be deduced. Changing the position of the theodolite station and adjusting the cross-screw to infinity, the transformation matrix between the base coordinate and star-sensor coordinate can be deduced by image formation of the destination at infinity. The position relation between object body and camera can be calculated by the transformation matrix. The results of the measurement experiment show that the gauging repeatability is 6' which can meet the system gauging demands.
Proc. SPIE. 7129, Seventh International Symposium on Instrumentation and Control Technology: Optoelectronic Technology and Instruments, Control Theory and Automation, and Space Exploration
KEYWORDS: Information fusion, Microelectromechanical systems, Error analysis, Complex systems, Star sensors, Gyroscopes, Velocity measurements, Algorithm development, Space operations, Systems modeling
Strapdown stellar-inertial integrated attitude determination based on low-cost micro-electromechanical system (MEMS)
gyroscopes and a complementary metal-oxide-semiconductor transistor active pixel star sensor is one of the most
effective methods for nano-spacecrafts attitude determination. However, the accuracy of attitude determination is low
because of non-linearity of the system. Thus, an adaptive segmented information fusion method based on the UKF is
presented by taking UKF+QUEST and UKF+optimal REQUEST as two modes of information fusion that can be
adaptively switched between. Initially, the gyro drift estimation error is inaccurate, and the UKF+QUEST mode is
adopted to quickly estimate the gyro drifts. When the mean-square error matrix of the UKF tends to stabilization, the
information fusion mode is adaptively switched to the UKF+optimal REQUEST dual-filter model. The hybrid simulation
results show this method not only has higher accuracy in attitude determination but also can quickly estimate gyro drifts.
Low-cost MEMS sensors often suffer from inaccuracy and are influenced greatly by temperature variation, and the
orientation error is cumulate with time. The GPS can provide long term stability with high accuracy, but it has its
insufficiencies, such as disturbed easy, lower data updating rate and so on, it is hard to meet the demand of real-time
measuring. The micro-magnetic sensors, an independence precision tool, can offer real-time yaw attitude angle, and this
can correct the orientation cumulate error, and it increase the independence in flight of the UAV. Based on analyzing the
selection principles of testing sensor, comparing 3 kinds micro-magnetic sensors, the GMI magnetic sensor is best to test
geomagnetic field. The Regional model of geomagnetic field is built, and a GMI-magnetic sensor navigation method is
put forward. Three-axis magnetic sensor measure the geomagnetic field, and it is matching with the geomagnetic map,
then the geomagnetic elements on currently position are knew, combining with the information of accelerometer, the
position information can be gotten by matching algorithm.