LabNet: An Image Repository for Virtual Science Laboratories - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2019

LabNet: An Image Repository for Virtual Science Laboratories

Ifeoluwatayo A. Ige
  • Function : Author
  • PersonId : 1054639
Bolanle F. Oladejo
  • Function : Author
  • PersonId : 1054640

Abstract

There has been recent research on image and shape storage and retrieval. Several image/shape repositories and databases of large datasets have existed in literature. However, it can be said that these repositories have generic image data content as most of them are English based images of the general world. Since they do not focus on specific field of interest while populating them, there is a high probability that they may not have a sufficient coverage for images and shapes related to specific domains or fields such as high school science-oriented images and shapes. Hence, we develop ‘LabNet’; an image repository for high school science which contains images of high school science-related subjects and laboratory courses. We use Canny’s algorithm for edge detection of objects from crawled images; and then perform morphological operation algorithms for segmentation and extraction of object images. We state that our object image does not have any background and can be utilized for scene modelling and synthesis. LabNet can also be useful for high school science-based research as well as an educational tool for elementary science-based classes and laboratory exercises.
Fichier principal
Vignette du fichier
484827_1_En_2_Chapter.pdf (816.03 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-02294705 , version 1 (23-09-2019)

Licence

Attribution

Identifiers

Cite

Ifeoluwatayo A. Ige, Bolanle F. Oladejo. LabNet: An Image Repository for Virtual Science Laboratories. International Working Conference on Transfer and Diffusion of IT (TDIT), Jun 2019, Accra, Ghana. pp.10-20, ⟨10.1007/978-3-030-20671-0_2⟩. ⟨hal-02294705⟩
55 View
22 Download

Altmetric

Share

Gmail Facebook X LinkedIn More