Real-time and Automatic Map Stitching through Aerial Images from UAV
Real-time aerial map stitching through aerial images had been done through many different methods. One of the famous methods was a features-based algorithm to detect features and match the features of two and more images to produce a map. There are several famous feature-based methods such as ORB, SIFT, SURF, KAZE, AKAZE and BRISK. These methods are the feature-based method to detect features and compute homography matrix from matches features to stitch images. There are many advantages to generate a real-time map from the viewpoints between 100-200m altitude. The high quality and fast generation of the map could help to save people from fire, flood, and other disasters. The ORB is the fastest and high precision method compared to another features-based method. However, the ORB method was used in C++ and python mostly, there is a lack of optimization of the structure to write the code in MATLAB. This project introduces to use matrix multiplication method to replace for loop to write the RANSAC algorithm in MATLAB software and change the workflow to detect the image features to increase the map stitching rate. The code was tested with an online aerial image dataset which contains 100 images with the resolution (640×480). The results achieve was an average 1.45 Hz update rate compared to the original code was 0.69 Hz. The method introduced in this thesis was successfully speeds up the processing time for the program to process map stitching.