Harald Schmitzer,
"Object Detection Using Hough Transform - Fuzzy Approach and Classical Methods"
, 6-1997, Harald Schmitzer. Object Detection Using Hough Transform - Fuzzy Approach and Classical Methods. Diplomarbeit, Johannes Kepler Universität Linz, Master thesis, 1997.
Original Titel:
Object Detection Using Hough Transform - Fuzzy Approach and Classical Methods
Sprache des Titels:
Englisch
Englische Kurzfassung:
Automation has been the main driving force of booming
industry and economy in this century. Before transistors and computers have been invented, automation was restricted to basic mechanical aspects only. With the help of computers and sensitive systems more complex and sophisticated tasks, which formerly were only able to be done by humans, could be managed by machines.
At the beginning there were only small systems that solved simple control problems. Later, when artificial intelligence has been developed and extended, the modelling of reality for computer algorithms became a major task. One special branch of this field is computer vision. Computer vision includes all problems that have to solved to realize world with some sensitive systems and to transform this information into describable objects that modell the aspects of reality that are needed for further
processing.
In this sense, object detection is defined as the extraction of certain classes of objects and features from a given visual input. Usually object detection can be divided into low-level and high-level-detection. For example low-level objects are simple features like edges, points, lines, circles or simple shapes. High-level objects are then built up from low-level objects and are
detected after calculating the low-level objects. Examples for high-level objects are complex geometric objects in 2D or 3D. This thesis deals with a special technique for low-level object detection called Hough Transform. The Hough Transform is a method that is able to detect simple objects like lines, but can also be extended to more general objects. The main idea on which the Hough Transform bases is that for each image feature the inverse
equation of the wanted object is solved. The solution is then a set of objects containing this feature. By calculating the intersection of these sets for all image features we get the object that is represented in the image.
Erscheinungsmonat:
6
Erscheinungsjahr:
1997
Notiz zum Zitat:
Harald Schmitzer. Object Detection Using Hough Transform - Fuzzy Approach and Classical Methods. Diplomarbeit, Johannes Kepler Universität Linz, Master thesis, 1997.
Forschungs-