TY - JOUR
T1 - AnalyZr: A Python application for zircon grain image segmentation and shape analysis
AU - Scharf, T.
AU - Kirkland, C. L.
AU - Daggitt, M. L.
AU - Barham, M.
AU - Puzyrev, V.
N1 - Funding Information:
This project was funded through Minerals Research Institute of Western Australia (MRIWA) grant M551 and Iluka Resources. The first author was funded by the Australian Government Research Training Program Stipend Scholarship. We thank the Geological Survey of Western Australia for providing extensive image datasets for this work.
Funding Information:
This project was funded through Minerals Research Institute of Western Australia (MRIWA) grant M551 and Iluka Resources . The first author was funded by the Australian Government Research Training Program Stipend Scholarship . We thank the Geological Survey of Western Australia for providing extensive image datasets for this work.
Publisher Copyright:
© 2022 The Authors
PY - 2022/5
Y1 - 2022/5
N2 - Zircon grain shape is traditionally interpreted as a product of the physico-chemical conditions during crystal growth and may be modified during grain transport processes. The analysis of magmatic zircon grain shape has been proposed to inform on crystallization conditions, whereas detrital zircon grain shape has been proposed to complement traditional sediment provenance analysis. Shape parameters can be automatically measured from digital images of zircon mounts; however, this requires extraction of individual grain boundaries for measurement. Existing image segmentation software may require the use of proprietary languages, or knowledge of scripting to develop automated image segmentation routines, and is typically not tailored towards the geosciences. Furthermore, the separation of touching zircon grains in images remains a challenge for existing algorithms. To facilitate zircon grain shape analysis, we present AnalyZr, an open-source graphical Python application designed to segment reflected and transmitted light images of zircons mounted in resin. A new segmentation algorithm is implemented to improve the separation of touching zircon grains. Shape parameters are automatically measured from the segmented images and may be output to a .csv or .mdb file. Two case studies demonstrate the use of the application in resolving geologically relevant information in zircon grains sourced from: i) compositionally and age-distinct granite, diorite, and gabbro samples from across Western Australia, and ii) age-distinct detrital zircons from the Canning Basin, Western Australia.
AB - Zircon grain shape is traditionally interpreted as a product of the physico-chemical conditions during crystal growth and may be modified during grain transport processes. The analysis of magmatic zircon grain shape has been proposed to inform on crystallization conditions, whereas detrital zircon grain shape has been proposed to complement traditional sediment provenance analysis. Shape parameters can be automatically measured from digital images of zircon mounts; however, this requires extraction of individual grain boundaries for measurement. Existing image segmentation software may require the use of proprietary languages, or knowledge of scripting to develop automated image segmentation routines, and is typically not tailored towards the geosciences. Furthermore, the separation of touching zircon grains in images remains a challenge for existing algorithms. To facilitate zircon grain shape analysis, we present AnalyZr, an open-source graphical Python application designed to segment reflected and transmitted light images of zircons mounted in resin. A new segmentation algorithm is implemented to improve the separation of touching zircon grains. Shape parameters are automatically measured from the segmented images and may be output to a .csv or .mdb file. Two case studies demonstrate the use of the application in resolving geologically relevant information in zircon grains sourced from: i) compositionally and age-distinct granite, diorite, and gabbro samples from across Western Australia, and ii) age-distinct detrital zircons from the Canning Basin, Western Australia.
KW - Boundary detection
KW - Grain separation
KW - Particle segmentation
KW - Shape measurement
KW - Zircon provenance analysis
UR - http://www.scopus.com/inward/record.url?scp=85125809418&partnerID=8YFLogxK
U2 - 10.1016/j.cageo.2022.105057
DO - 10.1016/j.cageo.2022.105057
M3 - Article
SN - 0098-3004
VL - 162
JO - Computers and Geosciences
JF - Computers and Geosciences
M1 - 105057
ER -