Skip to main content
ARS Home » Plains Area » Mandan, North Dakota » Northern Great Plains Research Laboratory » Research » Publications at this Location » Publication #353265
Title: Color calibration of digital images for agriculture and other applications

Author
item SHAJAHAN, SUNOJ - North Dakota State University
item CANNAYEN, IGATHINATHANE - North Dakota State University
item Saliendra, Nicanor
item Hendrickson, John
item Archer, David

Submitted to: Remote Sensing of Environment
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/14/2018
Publication Date: 10/21/2018
Citation: Shajahan, S., Cannayen, I., Saliendra, N.Z., Hendrickson, J.R., Archer, D.W. 2018. Color calibration of digital images for agriculture and other applications. Remote Sensing of Environment. 146:221-234. https://doi.org/10.1016/j.isprsjprs.2018.09.015.
DOI: https://doi.org/10.1016/j.isprsjprs.2018.09.015

Interpretive Summary: The growth phase and fitness of an agricultural crop could be assessed with digital photographs (images) through color changes. However, misleading inferences may arise due to variations in light conditions when images were taken. An image processing tool was developed, using ImageJ software, to obtain uniform color intensity (calibration) in all images using a standard color chart (24 patches). Image calibration was validated with actual images taken in laboratory and field conditions. Calibration performance using any 12 color patches, considered in any order, was as good as 24 colors; while commonly used neutral colors (white and grays) gave poor performance. We recommend three color patches, namely red (R), green (G) and blue (B), for calibration considering simplicity and practical usage. The tool needed ˜7 s to calibrate an image using a Windows laptop (Intel Core i5 and 8 GB RAM). Calibrated images from the developed tool, when used for determining crop growth phases (phenology) and other applications would be more reliable than simply using the raw images.

Technical Abstract: The image processing application in agriculture relies majorly on correlating color changes of images’ region of interest to determine quality attributes (e.g., plant phenology, plant health, crop stress, products ripeness). Changes in lighting conditions during image acquisition affect the image color, even though there is no change in the quality, and produces misleading inference when used without calibration. Developing a methodology to calibrate the images, so that they become homogeneous in intensity, for better phenological comparison is the focus of this study. The method was developed with synthetic images and validated with actual plant images in laboratory and field conditions using a standard X-Rite ColorChecker chart. Six different color schemes were tested to determine the effect of patch order, and minimum number of patches required for efficient calibration. A user-coded ImageJ plugin was developed in Fiji for color calibration that derived a [3 x 3] color calibration matrix, based on selected color patches and standard values. A new calibration performance index was developed to evaluate the calibration performance. Calibration using any 12 color patches taken in any order gave equal performance. Calibration performance using only commonly followed neutral color patches was poor. Using red (R), green (G), and blue (B) color patches was recommended as it produced visually similar images, the performance was comparable with 24 color patches, and was simple and practical. The developed plugin took ˜7 s for calibration (Windows laptop, Intel Core i5, and 8 GB RAM). Calibrated images from the plugin, when used for phenological and other applications comparison, would be more reliable than simply using the raw images.