Detekce beztexturových 3D objektů

Abstract

In this diploma thesis I deal with description and implementation of the algorithm for fast detection of textureless 3D objects in image. The target is not only detection, but also determine approximate position of the object. It is captured on the image by different sizes under different angles of view. Output of the algorithm is marked object with frame including positioning and value of probability to template.

Description

Subject(s)

openCV, detection of textureless objects, features, index table, cascade classifier, scanning window, OCM, non-maxima suppression

Citation