Complex Object Detection Using Light-Field Plenoptic Camera

Edgar S. Correa, Carlos A. Parra, Pedro R. Vizcaya, Francisco Carlos Calderon, Julian D. Colorado

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

Identifying objects with a lot of complex features in an uncontrolled scene is a challenging task in the image processing research field. This article presents the application of robust and invariant feature descriptors together with an optimal, iterative and probabilistic methodology. These methods let to achieve an application of object detection through a homographic transformation matrix. This work is built on the Raytrix R42 Plenoptic camera, for which the camera calibration process is introduced. The results present a homographic transformation that relates spatial information of the reference object in an uncontrolled scene. This approach put forward an efficient performance that overcomes bad lighting and occlusion problems.

Original languageEnglish
Title of host publicationRecent Trends in Image Processing and Pattern Recognition - 4th International Conference, RTIP2R 2021, Revised Selected Papers
EditorsKC Santosh, Ravindra Hegadi, Umapada Pal
PublisherSpringer Science and Business Media Deutschland GmbH
Pages119-133
Number of pages15
ISBN (Print)9783031070044
DOIs
StatePublished - 2022
Event4th International Conference on Recent Trends in Image Processing and Pattern Recognition, RTIP2R 2021 - Virtual, Online
Duration: 08 Dec 202110 Dec 2021

Publication series

NameCommunications in Computer and Information Science
Volume1576 CCIS
ISSN (Print)1865-0929
ISSN (Electronic)1865-0937

Conference

Conference4th International Conference on Recent Trends in Image Processing and Pattern Recognition, RTIP2R 2021
CityVirtual, Online
Period08/12/2110/12/21

Keywords

  • Homography transform
  • Optimal statistic approach
  • Plenoptic camera
  • Robust descriptor

Fingerprint

Dive into the research topics of 'Complex Object Detection Using Light-Field Plenoptic Camera'. Together they form a unique fingerprint.

Cite this