Data from: Automated size measurements of Halyomorpha halys (Stål)(Heteroptera: Pentatomidae) with simple image-based methodology

This data is a companion to this paper: 

Amy Tabb, Johanna E Elsensohn, Tracy C Leskey, Automated size measurements of Halyomorpha halys (Stål)(Heteroptera: Pentatomidae) with simple image-based methodology," 2022, Florida Entomologist, 105(3):262-264. doi: 10.1653/024.105.0314 

@article{tabb_automated_2022,author = {Amy Tabb and Johanna E. Elsensohn and Tracy C. Leskey},title = {{Automated Size Measurements of Halyomorpha halys (Stål) (Heteroptera: Pentatomidae) with Simple Image-Based Methodology}},volume = {105},journal = {Florida Entomologist},number = {3},publisher = {Florida Entomological Society},pages = {262 -- 264},keywords = {brown marmorated stink bug, chinche apestosa marrón marmorada, computer vision, morfología, morphology, visión por computador},year = {2022},doi = {10.1653/024.105.0314},URL = {https://doi.org/10.1653/024.105.0314}}

The paper makes use of a tool called CASS. We list its information here as well:

A. Tabb, Germán A. Holguín, and Rachel Naegele, “Using Cameras for Precise Measurement of Two-Dimensional Plant Features: CASS,” in High-Throughput Plant Phenotyping: Methods and Protocols, A. Lorence and K. Medina Jimenez, Eds. New York, NY: Springer US, 2022, pp. 87–94. doi: 10.1007/978-1-0716-2537-8_10. or https://arxiv.org/abs/1904.13187

Manually-measured data

The ground-truth data is in Dead-BMSB-measurements-ground-truth.xlsx.

Reproducing graphs

The directory plots has the following structure:plots- dataLength.dat  - gt_bottomLength.dat  - plottingOctave.m- dataWidth.dat- gt_width.txt

To recreate the graphs from the paper, run script plottingOctave in GNU Octave. It is likely to run as well in Matlab, but we have not tested. We have used Octave vers. 4.2.1. dataLength.dat and dataWidth.dat are the results from the automated computer vision method.

Data structure

Raw image data of Brown Marmorated Stink Bug (BMSB) on calibration patterns is in input-images\images. The full structure of the input directory is as follows:

input-images- calibration_object_info.txt- exif-log.txt- sensor_size.txt- created_template.png- specification_file.txt- images (directory)

calibration_object_info.txt : Gives the size dimensions of the aruco pattern.

exif-log.txt : Exif information from one of the images.

sensor_size.txt : Image sensor dimensions of the iPhone 8, needed for CASS to perform calibration from a single image.

created_template.png : The calibration pattern created by CASS; in the case of this project, the center of the pattern was masked and only the border patterns remain.

specification_file.txt : Information concerning the aruco patterns, needed for CASS to perform calibration from a single image.

 images : The first three characters are the number of the insect, i.e. 001 is 1, 100 is 100, etc. All are images from an iPhone 8.

Processed images structureThe processed images are in the directory result-images. Directory cass has the result of applying the single-image calibration and warping, such that every 10 pixels is equivalent to 1 mm, and cropped out from the surrounding background calibration patterns. The segmentation directory images have the insect segmented from the background, using the warped images from the cass directory.