Algorithm of Ground Object Detection with Modified SURF Method by Using Morphological Features

摘要

Two approaches are combined into one algorithm in order to reach better accuracy of ground object detection and tracking. Blob analysis provides the separation of detected object by morphological criterion (mostly minimum area threshold is used) and is used to determine the region of interest for further selection of feature points. Points are detected and described by Speed-Up Robust Feature descriptors method. Due to combination the amount of processing information is decrease that leads to increasing of not only the accuracy but also efficiency of algorithm.