https://scholars.lib.ntu.edu.tw/handle/123456789/633901
Title: | Multi-Stage Superpixel-Based Segmentation Algorithm Using Fully Convolutional Networks and Discriminative Features | Authors: | Huang, Pei Chi JIAN-JIUN DING |
Issue Date: | 1-Jan-2022 | Source: | Proceedings of 2022 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2022 | Abstract: | Image segmentation is a process to partition an image into multiple segments. Pixels within the same segments share certain characteristics. Therefore, in this paper, we simplify the segmentation task to a superpixel-merging decision problem by considering discriminative features. Characteristics, including border, color, brightness, area size, sharpness, saliency, texton, and etc. are taken into account. We propose a multi-stage superpixel-clustering algorithm to merge superpixels of similar characteristics. In the first stage, the fully convolutional network is applied to decide whether the boundary of two adjacent superpixels should be kept or not. The model is trained basing on color, superpixel boundary, and edge detection result of the image. In the second stage, scoring method and SVM classification model are used to further decide whether the rest superpixels should be merged or not. We consider up to 14 factors to further improve the performance. Overall, simulations and evaluation metrics show that our algorithm has highly accurate segmentation results. |
URI: | https://scholars.lib.ntu.edu.tw/handle/123456789/633901 | ISBN: | 9786165904773 | DOI: | 10.23919/APSIPAASC55919.2022.9979963 |
Appears in Collections: | 電機工程學系 |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.