Recognition of two different kinds of objects using machine vision: Saimaa ringed seals and traffic signs

ABGESAGT: Vortrag im Kolloquium Informatik von Prof. Kälviäinen

Der Termin fällt leider aus!



Mittwoch, 19.04.2018, 16:15 Uhr, Raum G 309


Referent: Prof. Heikki Kälviäinen  (Lappeenranta University of Technology)

Gastgeber: Prof. Dr. Dietrich Paulus


Titel: Recognition of two different kinds of objects using machine vision: Saimaa ringed seals and traffic signs



Research at Machine Vision and Pattern Recognition Laboratory (MVPR, from Lappeenranta University of Technology (LUT) focuses on applications of machine vision and digital image processing. Two different application areas are considered in the presentation: identification of Saimaan ringed seals and condition analysis of traffic signs. The first application area is related to conservation of nature. In order to monitor animal. population and to track individual animals in a non-invasive way, identification of individual animals based on certain distinctive characteristics is necessary. In this presentation, automatic image-based individual identification of the critically endangered Saimaa ringed seal (Phoca hispida saimensis) is considered. The ringed seals have a distinctive permanent pelage pattern that is unique to each individual. A framework is proposed, including the segmentation of the seal from the background, the post-processing of the seal image for better recognition, and finally the identification based on machine learning. For experiments, the novel dataset of the Saimaa ringed seal images has been collected and annotated. The second application area is about automatic traffic sign inventory and simultaneous condition analysis. This machine vision application can be used to improve road maintenance processes, to decrease maintenance costs, and to meet the requirements of future intelligent driving systems. The goal of this research is to combine automatic traffic sign detection and classification with traffic sign inventory and condition analysis. The performance of the system has been evaluatedwith three datasets: the Swedish summer dataset and two datasetsspecifically collected for this research.

Wann 19.04.2018
von 16:15 bis 17:45
Wo G 309
Name Prof. Dr. Dietrich Paulus
Termin übernehmen vCal