Semi-Automatic Image Labelling Using Depth Information
Image labeling tools help to extract objects within images to be used as ground truth for learning and testing in object detection processes. The inputs for such tools are usually RGB images. However with new widely available low-cost sensors like Microsoft Kinect it is possible to use depth images...
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2015-05-01
|
Series: | Computers |
Subjects: | |
Online Access: | http://www.mdpi.com/2073-431X/4/2/142 |
_version_ | 1811183911713112064 |
---|---|
author | Mostafa Pordel Thomas Hellström |
author_facet | Mostafa Pordel Thomas Hellström |
author_sort | Mostafa Pordel |
collection | DOAJ |
description | Image labeling tools help to extract objects within images to be used as ground truth for learning and testing in object detection processes. The inputs for such tools are usually RGB images. However with new widely available low-cost sensors like Microsoft Kinect it is possible to use depth images in addition to RGB images. Despite many existing powerful tools for image labeling, there is a need for RGB-depth adapted tools. We present a new interactive labeling tool that partially automates image labeling, with two major contributions. First, the method extends the concept of image segmentation from RGB to RGB-depth using Fuzzy C-Means clustering, connected component labeling and superpixels, and generates bounding pixels to extract the desired objects. Second, it minimizes the interaction time needed for object extraction by doing an efficient segmentation in RGB-depth space. Very few clicks are needed for the entire procedure compared to existing, tools. When the desired object is the closest object to the camera, which is often the case in robotics applications, no clicks at all are required to accurately extract the object. |
first_indexed | 2024-04-11T13:04:49Z |
format | Article |
id | doaj.art-36fe167aa1374465ae9e33bfa13aa8f7 |
institution | Directory Open Access Journal |
issn | 2073-431X |
language | English |
last_indexed | 2024-04-11T13:04:49Z |
publishDate | 2015-05-01 |
publisher | MDPI AG |
record_format | Article |
series | Computers |
spelling | doaj.art-36fe167aa1374465ae9e33bfa13aa8f72022-12-22T04:22:49ZengMDPI AGComputers2073-431X2015-05-014214215410.3390/computers4020142computers4020142Semi-Automatic Image Labelling Using Depth InformationMostafa Pordel0Thomas Hellström1Department of Computing Science, Umeå University, Umeå, SE-901 87, SwedenDepartment of Computing Science, Umeå University, Umeå, SE-901 87, SwedenImage labeling tools help to extract objects within images to be used as ground truth for learning and testing in object detection processes. The inputs for such tools are usually RGB images. However with new widely available low-cost sensors like Microsoft Kinect it is possible to use depth images in addition to RGB images. Despite many existing powerful tools for image labeling, there is a need for RGB-depth adapted tools. We present a new interactive labeling tool that partially automates image labeling, with two major contributions. First, the method extends the concept of image segmentation from RGB to RGB-depth using Fuzzy C-Means clustering, connected component labeling and superpixels, and generates bounding pixels to extract the desired objects. Second, it minimizes the interaction time needed for object extraction by doing an efficient segmentation in RGB-depth space. Very few clicks are needed for the entire procedure compared to existing, tools. When the desired object is the closest object to the camera, which is often the case in robotics applications, no clicks at all are required to accurately extract the object.http://www.mdpi.com/2073-431X/4/2/142image labellingimage segmentationdepth informationlabelling toolsRGBD dataMicrosoft Kinectobject detectionrobot vision |
spellingShingle | Mostafa Pordel Thomas Hellström Semi-Automatic Image Labelling Using Depth Information Computers image labelling image segmentation depth information labelling tools RGBD data Microsoft Kinect object detection robot vision |
title | Semi-Automatic Image Labelling Using Depth Information |
title_full | Semi-Automatic Image Labelling Using Depth Information |
title_fullStr | Semi-Automatic Image Labelling Using Depth Information |
title_full_unstemmed | Semi-Automatic Image Labelling Using Depth Information |
title_short | Semi-Automatic Image Labelling Using Depth Information |
title_sort | semi automatic image labelling using depth information |
topic | image labelling image segmentation depth information labelling tools RGBD data Microsoft Kinect object detection robot vision |
url | http://www.mdpi.com/2073-431X/4/2/142 |
work_keys_str_mv | AT mostafapordel semiautomaticimagelabellingusingdepthinformation AT thomashellstrom semiautomaticimagelabellingusingdepthinformation |