Apply Sequential Adjustment In The Analytical Aerotriangulation

During of data processing in the airborne remote sensing, the rotation angle is not too much big. We can consider the small angle in the adjustment. In addition; we can receive big images and many accurate pixel coordinates by using the measurement camera. Although it has these advantages, it can’t meet the people’s needs because of the high price and huge volume. With the development of UAV technology and non-metric digital camera, the low-altitude remote sensing becomes more realistic with the lower cost. This page do the further research based on the traditional remote sensing to solve the problem of angle and many points. The main contents are as follows:1) People need to find some standard points in the traditional orientation, while in the UAV, we can match a considerable amount of points and hard to find points like traditional orientation. If people consider all of them to solve the unknown parameters, it will result in difficulties in the matrix inversion proccessing.so we divide the points into several groups to adjustment. The result showed that it can get the similar precision with the traditional adjustment.2) This page gets the new relative orientation, absolute orientation and bundle adjustment by using Unit Quaternions instead of the Euler angles. Then it can avoid the trigonometric problem. The result showed that it can simplify the error equation and get faster convergence.3) To take account of the reliability, we select the sequential adjustment with additional parameters and change the rotational matrix. It’s easy to solve the problem of a large number of points and the mass equation. At the same time, we compared two methods. The result showed that the accuracy may not be high when we get the higher reliability. Therefore, to make a choice which one would be better, we should consider all aspects of them.4) In the processing of the bundle adjustment, it will produce a large-scale matrix due to applying a large number of points. To solve these problems; first, we can decompose the matrix directly by using the method of SVD. This method can solve the problem of the matrix singular. Second, we can do the successive adjustment to every image based on the sequential algorithm with a new parameter which can simplify the size of the matrix.