Image Filtering and Labelling Assistant (IFLA): Expediting the analysis of data obtained from camera traps

Field monitoring projects consistently generate a large volume of captured images. Biology/ecology researchers must sift out the useful images (i.e., those that contain animals) and use their expertise to label them prior to analysis, which is a laborious task when performed manually. In this study,...

Full description

Saved in:
Bibliographic Details
Published in:Ecological informatics 2021-09, Vol.64, p.101355, Article 101355
Main Authors: Xi, Tianyu, Wang, Jiangning, Qiao, Huijie, Lin, Congtian, Ji, Liqiang
Format: Article
Language:English
Subjects:
Biodiversity monitoring
Camera traps
CForBio
Empty image
Image processing
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Field monitoring projects consistently generate a large volume of captured images. Biology/ecology researchers must sift out the useful images (i.e., those that contain animals) and use their expertise to label them prior to analysis, which is a laborious task when performed manually. In this study, we developed an Image Filtering and Labelling Assistant (IFLA) system to expedite the most time-consuming portion of this process. This system supplies object-marked images to help researchers identify and label those that are useful. Initial evaluations showed that IFLA is more exact than manual methods. We also implemented an automated method for image selection and labelling, though its stability needs improvement. Tests show that IFLA can help volunteers save 30% of their time and improve 30% accurate in labelling images. •In this paper, we developed an IFLA (Image Filtering and Labeling Assistant) system to speed up the most time-consuming work of detecting useful images.•IFLA supplies the object-marked images to help researchers select the useful images and label them.•Initial evaluations show that this method is more exact than manual methods. IFLA also implements an automated method for selection and labeling
ISSN:1574-9541
DOI:10.1016/j.ecoinf.2021.101355