In:
Biodiversity Data Journal, Pensoft Publishers, Vol. 8 ( 2020-12-10)
Abstract:
As herbarium specimens are increasingly becoming digitised and accessible in online repositories, advanced computer vision techniques are being used to extract information from them. The presence of certain plant organs on herbarium sheets is useful information in various scientific contexts and automatic recognition of these organs will help mobilise such information. In our study, we use deep learning to detect plant organs on digitised herbarium specimens with Faster R-CNN. For our experiment, we manually annotated hundreds of herbarium scans with thousands of bounding boxes for six types of plant organs and used them for training and evaluating the plant organ detection model. The model worked particularly well on leaves and stems, while flowers were also present in large numbers in the sheets, but were not equally well recognised.
Type of Medium:
Online Resource
ISSN:
1314-2828
,
1314-2836
DOI:
10.3897/BDJ.8.e57090
DOI:
10.3897/BDJ.8.e57090.figure1
DOI:
10.3897/BDJ.8.e57090.figure2
DOI:
10.3897/BDJ.8.e57090.figure3
DOI:
10.3897/BDJ.8.e57090.figure4
DOI:
10.3897/BDJ.8.e57090.figure5
DOI:
10.3897/BDJ.8.e57090.figure6a
DOI:
10.3897/BDJ.8.e57090.figure6b
DOI:
10.3897/BDJ.8.e57090.figure6c
DOI:
10.3897/BDJ.8.e57090.figure6d
DOI:
10.3897/BDJ.8.e57090.suppl1
DOI:
10.3897/BDJ.8.e57090.suppl2
DOI:
10.3897/BDJ.8.e57090.suppl3
Language:
Unknown
Publisher:
Pensoft Publishers
Publication Date:
2020
detail.hit.zdb_id:
2736709-5