Automatic Construction of Hypotheses for Linear Objects in Digital and Laser Scanning Images

Abstract

This paper presents an automatic road hypotheses approach using digital image and laser scanning image combining various Digital Image Processing techniques. The semantic objects, in this work, are linear features, such as, roads and streets. The aim of this paper is extract automatically road hypotheses on image space and object space for use the information in automatic absolute orientation process. The results show that methodology is efficiency and the roads hypotheses are generate and validate.

Keywords

• Automatic construction of road hypotheses
• Feature extraction straight lines
• Semantic objects
• Image classification.