Browsing by Author "Cetinkaya,B."

Now showing 1 - 3 of 3

A Component-Based Object Detection Method Extended With a Fuzzy Inference Engine
(Institute of Electrical and Electronics Engineers Inc., 2015) Koyuncu,M.; Cetinkaya,B.
In this paper, we propose a component-based object detection method extended with the fuzzy inference technique. The proposed method detects constituent components of a complex object instead of a whole object in images. For component detection, multiple multi-class support vector machines (SVM) are used in parallel. Each SVM classifies the candidate component using a different low-level image feature. The obtained results are fused to reach a decision about the component. Then, a fuzzy object extractor determines the whole object considering the detected components and their geometric configurations. The fuzzy object extractor is a fuzzy inference engine which tests various combinations of detected components and their fuzzified directions and distances. The initial tests yield promising results and encourage further studies to extend proposed method. © 2015 IEEE.
Feature fusion in part-based object detection;
(Institute of Electrical and Electronics Engineers Inc., 2015) Koyuncu,M.; Cetinkaya,B.
In this study, classification of complex objects in images as a whole is compared with classification of its distinctive components using different features. In addition, the impact of feature fusion in part-based object detection is investigated. Applied method, implemented system, conducted tests and their results are presented in this paper. Test results show that, even in the case of a good segmentation, object components are classified more successfully compared to whole object and feature fusion method improves the obtained results to a certain degree. © 2015 IEEE.
Part-Based Object Extraction for Complex Objects;
(IEEE Computer Society, 2014) Koyuncu,M.; Cetinkaya,B.
Indexing requirement for efficient accessing to visual data has been increased with the widespread use of multimedia applications. Satisfaction of this requirement mostly depends on the automatic extraction of objects in the visual data. In this study, component-based object extraction method is compared with object extraction in its entirety. Applied method, implemented system and conducted tests are presented in this paper. Test results show that, even in the case of a good segmentation is achieved for whole object, object components are classified more successfully compared to whole object. © 2014 IEEE.