Title :
Ensemble multisensor data using state-of-the-art classification methods
Author :
Twala, Bhekisipho ; Mekuria, Fisseha
Author_Institution :
Dept. of Electr. & Electron. Eng. Sci., Univ. of Johannesburg, Johannesburg, South Africa
Abstract :
Detection and identification from sensing image is a common task for many applications. In order to improve the performance of detection and identification the use of multiple classifier combination is demonstrated and evaluated in the paper using two industrial image datasets. Experiments show that multiple classifier combination can improve the performance of image classification and image detection and identification with boosting and bagging achieve higher accuracy rates. Accordingly, good performance is consistently derived from static parallel systems.
Keywords :
image classification; image fusion; learning (artificial intelligence); object detection; accuracy rates; bagging; boosting; classification methods; ensemble multisensor data; image classification; image detection; image identification; industrial image datasets; multiple classifier combination; sensing image; static parallel systems; Accuracy; Artificial neural networks; Bagging; Boosting; Error analysis; Logistics; Training; ensemble systems; machine learning; multiple classifiers; sensor data; target tracking;
Conference_Titel :
AFRICON, 2013
Conference_Location :
Pointe-Aux-Piments
Print_ISBN :
978-1-4673-5940-5
DOI :
10.1109/AFRCON.2013.6757711