Sunflower floral dimension measurements using digital image processing

• Published in: Computers and electronics in agriculture 2018-08, Vol.151, p.403-415
• Main Authors: Sunoj, S., Subhashree, S.N., Dharani, S., Igathinathane, C., Franco, J.G., Mallinger, R.E., Prasifka, J.R., Archer, D.
• Format: Article
• Language: English
• Subjects: Algorithm; Algorithms; Digital imaging; Dimension measurement; Flowers & plants; Image acquisition; Image processing; Image processing systems; Measurement methods; Operating systems; Plant reproduction; Pollinator; Prediction models; Projective transform
• ISSN: 0168-1699; 1872-7107
• DOI: 10.1016/j.compag.2018.06.026

[Display omitted] •An ImageJ plugin for sunflower head, disc, and ray florets dimensions was developed.•Plugin automatically rectifies image distortions and performs multiple measurements.•At least 32 measurements required for sunflower head and ray florets, and 8 for disc.•Plugin serves as a tool for entomologists for studying pollinators interaction. Sunflower floral dimensions are essential for assessing pollinators attraction and estimating seed yields. Dimensions measured manually at present are subjective and time-intensive, therefore, an image processing method was developed as an alternative, which was objective, non-destructive, produces various outputs, and rapid. An ImageJ user-coded plugin with a field image acquisition method was developed to measure the dimensions of individual sunflower components, such as head, disc, and ray florets. Two measurement methods, direct (using the thresholded binary image) and wrapping-polygon (using a polygonal enclosure) were tested. The ‘pixel-march’ method made multiple radial dimension measurements (diameter) on the ray florets binary image in a single computation. The effect of multiple measurements (2, 4, 8, 16, 32, 64, 128, and 180 along 0–180° angles) was studied to determine an effective number of measurements, and user-friendly sunflower dimension prediction models from ImageJ’s standard output parameters were developed. Results indicated that (i) a minimum of 32 measurements for the sunflower head and ray florets dimensions, but only eight measurements for the sunflower disc, were necessary; (ii) wrapping-polygon method was efficient compared to direct; (iii) equivalent diameter (ED) and fitted ellipse minor axis (MinA) were well correlated (r≥0.88) to the accurate mean 180 measurements (D180) for all sunflower components; (iv) linear models for predicting D180 using ED and MinA performed better (R2>0.99) for head and disc than for ray florets (R2>0.76); (v) user-friendly linear models using the mean of two manual measurements of the head (D2h) for predicting D180 and area were good only for the head (R2>0.92), and not suitable for disc (R2≤0.62) and ray florets (R2≤0.43); and (vi) the developed image processing method results were accurate, quick (≈11 s in Windows 10, Intel Core i5, and 8 GB RAM laptop), and have the potential to be adapted to other species.