Study On Method Of Self-reference Scanning Measurement Based On Inertial Navigation

In order to solve the problem of 3-D shape measurement of large size parts,a self locating 3-D scanning measurement system based on inertial navigation was built,which realized fast scanning of three-dimensional objects.The inertial measurement unit was integrated into the binocular vision measurement system in our designed system.By using the data fusion technology,the function of self-localization and 3-D shape information acquisition was realized.To achieve self-localization of the measurement system,the 3-D contour information of the object that measured by binocular vision measurement system was needed to be converted to a same coordinate system.In this paper,we used the coordinate system of the first position that measured by our vision measurement system as the ultimate target coordinate system.Before measurements,we need to paste certain circular markers made of high reflectivity material on the surface of the object to be measured,and then calibrated the system parameters.During the process of measuring,the binocular vision measurement system measured the feature points and the 3-D coordinate of structured light which included the object contour information simultaneously.Then based on the feature points,we derived the coordinate system transformation relationship of the vision measurement system under different positions.By conducting data fusion with data measured by the inertial measurement unit,more accurate location information could be determined.Above all,the coordinate values of all the data measured by vision measurement system could be unified.The main content of this paper is as follows:1?The working principle of self-locating 3-D scanning measurement based on inertial navigation and the calibration of the system.This paper presented an implement method for detecting the position of the measurement system by combining an inertial navigation with vision measurement system.Its realization process was introduced in detail and the relevant verification experiment was designed.Then we designed an experimental scheme that was suitable for the system calibration.This experimental scheme can finish the calibration of binocular camera,the calibration of linear structured light plane and the calibration of the position relation between inertial measurement unit and binocular vision system in one-time.2?Research on data splicing method based on feature points.The feature points measured by binocular vision measurement system under different angle of view need to be correctly matched to ensure the accuracy of subsequent data splicing.This paper realized the matching of same name tag points by constructing congruent triangles.On this basis,the transformation matrix was acquired using the method of solving unit quaternion.Then the measuring data splicing was achieved.3?Research on the fusion of the data that measured by inertial measurement unit and vision measurement system.On the basis that the binocular vision measurement system used the feature points to derive the coordinate transformation relationship under different angle of view,we used the proposed data fusion method to calculate the more accurate coordinate transformation matrix.Then self-location of the measurement system and data splicing could be achieved.