Search

This paper addresses the problem of autonomous navigation of UGV in an unstructured environment. Generally, state-of-the-art approaches use color based segmentation of road/non-road regions in particular. There arises an important question, how is the distance between an input pixel and a color model measured. Many algorithms employ Mahalanobis distance, since Mahalanobis distance better follows the data distribution, however it is assumed, that the data points have a normal distribution. Recently proposed Polynomial Mahalanobis Distance (PMD) represents more discriminative metric, which provides superior results in an unstructured terrain, especially, if the road is barely visible even for humans. In this paper, we discuss properties of the Polynomial Mahalanobis Distance, and propose a novel framework - A Three Stage Algorithm (TSA), which deals with both, picking of suitable data points from the training area as well as self-supervised learning algorithm for long-term road representation. [ABSTRACT FROM PUBLISHER]

This paper deals with new ideas and progress made in the development of a software environment for 3D computations. As the previous versions it is capable of camera intrinsic and extrinsic parameters calibration and 3D point coordinates reconstruction. In addition to that new features have been implemented including linearity and planarity constraints together with lens distortion identification. [ABSTRACT FROM AUTHOR]

This paper describes the design philosophy for reconstruction software. The main purpose of the software is 3D reconstruction from images. The program is able to solve the main problems from image processing and 3D reconstruction of objects. The software has been designed so that every possible task is solved in one program with one mathematical engine. Contrary to older versions of the software greater attention is applied to the modular structure of the program that has led to the expansion of the program potentials, especially in the addition of new objects and their relations. The program not only works with objects and their projections, but is also able to work with relations between objects. [ABSTRACT FROM AUTHOR]

This paper describes multipurpose software for reconstruction 3D shapes from 2D images. This software tool is used especially for points and lines (distances). It can count and measure 3D coordinates and distances for points and lines. The result is 3D coordinates or 3D wire model if needed. The main goal of the solution is to design software which can solve different types of tasks in one program unit with the same counting core and simple data sharing - result of one calculus can be used as input for another counting in the same program. [ABSTRACT FROM AUTHOR]

This paper describes visual system for non-contact inspection of glass bottles. The main aim is to detect flaws for rejection of faulty bottles from industrial flow line automatically. Flaws detection and localization is done by means of complex camera system with special image processing algorithms. Defects as a dirt, crack, mildew, remainders of labels etc. form inhomogeneous regions on uniform glass material. These distinctive non-homogeneities are detected in all images of the bottom, neck and side of the each bottle. In addition to mentioned basic image processing methods there is also included pattern recognition sub-system, which is able to distinguish between synthetic patterns (e.g. manufacturer logo) and genuine flaws (e.g. mildew in the bottle). [ABSTRACT FROM AUTHOR]