ABSTRACT
ABSTRACT The geometric distortion in push-broom digital aerial imagery can be rectified using data from an inertial measurement unit (IMU). However, inaccuracies in the IMU data cause undulant, wavelike twist deformations in the push-broom digital aerial images after geometric calibration, which directly influences the authenticity and liability of the images and their practical application. At present, the diagnosis of image deformation depends mainly on the subjective judgment of a human being, which costs a great deal of time and manpower. In this chapter, an automatic method
CONTENTS
12.1 Introduction ................................................................................................ 198 12.2 Method of Inspection for Deformation in Digital Aerial Imagery
Based on Statistical Characteristics .........................................................200 12.2.1 Contour Extraction and Preprocessing of Digital Aerial
Imagery ........................................................................................... 201 12.2.1.1 Edge Detection ................................................................ 202 12.2.1.2 Contour Tracing .............................................................. 203 12.2.1.3 Curve Fitting .................................................................... 204
12.2.2 Search for and Judgment of Squiggles in Aerial Imagery ....... 204 12.2.2.1 Search for Extreme Points of Curvature ...................... 204 12.2.2.2 Judgment of Squiggles .................................................... 206
12.2.3 Statistical Directions and Distortion Judgment for the Squiggles of Aerial Imagery ......................................................... 207 12.2.3.1 Statistical Directions of Squiggles ................................ 207 12.2.3.2 Judgment of Deformations ............................................ 208
12.3 Experiments and Results .......................................................................... 209 12.4 Summary ..................................................................................................... 210 References ............................................................................................................. 210
of inspection for deformation in digital aerial imagery based on statistical characteristics is proposed to inspect for the distortions caused by inaccurate IMU data. Because the undulant, wavelike deformations in an image regularly feature the displacement of pixels in the same direction, many wave curves in the image appear in the same direction after geometric calibration. Therefore, in this method, the positions of the wave curves in the image are located by the extreme points of curvature of the contour lines, and the wavelike deformations can then be judged automatically through the distribution statistics of the open directions of the wave curves. The specific method implemented can be described as follows. First, the edges of the image are detected with a Canny edge detector, and the vector contour lines are obtained by tracing the edges and fitting them with the cubic spline curve method. The extreme points of curvature of the contour lines are then calculated, and some of these points are determined to be the vertices of the wave curves by judging the positional relationships between each extreme point and the points around them, thus constituting a vertex set. The perpendicular directions of the tangents of the vertices are then used as the directions of the wave curves, and the direction histograms of all of the wave curves in the image are obtained by statistical analysis. Finally, the existence of the deformation phenomenon in the image due to the inaccurate IMU data can be determined on the basis of whether the directions of the wave curves are centralized in a certain direction. The experimental results show that the automatic method of inspection for deformation presented in this chapter can effectively detect with 95% accuracy the deformation in digital aerial images caused by inaccurate IMU data.
