Data and Code from: Using cameras for precise measurement of two-dimensional plant features: CASS

Computer vision explanation: The code (https://github.com/amy-tabb/CASS, referred to as CASS) takes an image of an object on top of an aruco calibration pattern, calibrates the camera using the detected aruco information as well as EXIF tag information, and undistorts and computes the homography from the current location of the aruco calibration pattern in the image to its location in physical space. Then the image is warped to match the coordinate system of the aruco coordinate system, scaled by a user-selected parameter.

This dataset provides examples of properly-formatted input images and accompanying text files, as well as a successful run where the option of writing intermediate results has been selected. Details about how to format the directories is found in the README of https://github.com/amy-tabb/CASS.

• iphone6. is a directory of input files using the camera of a iPhone 6 cellular phone.
• iphone6_results. is the directory of results created from running CASS on iphone6.
• CanonEOS60D. is a directory of input files using a DSLR camera from Canon, model name EOS 60D.
• CanonEOS60D_results. is the directory of results created from running CASS on CanonEOS60D.

See the Write directory format section of CASS's README for details of all of the files; briefly for this example, warped_ORIGINALFILENAME.jpg is the original image, transformed such that 10 pixels corresponds to 1 millimeter on the two dimensional plane of the calibration pattern.

https://github.com/amy-tabb/CASS provides code in C++ for processing on this dataset, as well as as Docker image.