| |
Web URL(s): | https://archive.lib.msu.edu/tic/its/articles/2009jou437.pdf Last checked: 10/18/2011 Requires: PDF Reader |
Access Restriction: | Certain MSU-hosted archive URLs may be restricted to legacy database members. |
Publication Type:
| Refereed |
Author(s): | Narra, Siddhartha;
Fermanian, Thomas W.;
Voigt, Thomas B. |
Author Affiliation: | Department of Natural Resources and Environmental Sciences, University of Illinois at Urbana-Champaign, Urbana, Illinois |
Title: | A machine vision approach to objective turfgrass texture evaluation |
Section: | Establishment and maintenance Other records with the "Establishment and maintenance" Section
|
Meeting Info.: | Santiago, Chile: July 26-30 2009 |
Source: | International Turfgrass Society Research Journal. Vol. 11, No. Part 1, 2009, p. 437-448. |
Publishing Information: | Madison, WI: International Turfgrass Society |
# of Pages: | 12 |
Keywords: | TIC Keywords: Image analysis; Quality evaluation; Texture
|
Abstract/Contents: | "Past research has identified significant intra-rater and inter-rater variability in turfgrass texture evaluations using traditional quality rating systems. The objectives of this study were to overcome the drawbacks of the traditional evaluation method through a machine vision approach. A global run-length encoding (RLE) algorithm was initially developed and verified on digital images collected from a simulated turf environment built using commercially available twist ties. Regression equations developed from manual measurements of twist ties with different widths, lengths and densities, and the algorithm-derived widths showed significant relationship with R2 values of 0.62 and 0.74 for planar and turf-type twist-tie arrangements, respectively. The RLE algorithm was later applied to canopy images collected from an unofficial National Turfgrass Evaluation Program (NTEP) trial and Cooperative Turfgrass Breeders Test (CTBT) trial of Kentucky bluegrass. The regression equation generated from the turf-type arrangement of twist ties was used to predict texture from canopy images of different cultivars. The implemented RLE technique effectively quantified leaf blade width from cultivars with a range of texture characteristics. The algorithm was most effective in the case of non-touching leaves, while the correlations were less significant in images where the leaves were either overlapping or occluding. Significant correlations (-0.58 to -0.71 for different studies) were also present between subjective ratings and algorithm-derived values. Inter-plot and intra-plot variation was significantly higher than intra-image variability for both NTEP and CTBT trial cultivars." |
Language: | English |
References: | 15 |
Note: | Pictures, b/w Figures Tables Graphs |
| ASA/CSSA/SSSA Citation (Crop Science-Like - may be incomplete): Narra, S., T. W. Fermanian, and T. B. Voigt. 2009. A machine vision approach to objective turfgrass texture evaluation. Int. Turfgrass Soc. Res. J. 11(Part 1):p. 437-448. |
| Fastlink to access this record outside TGIF: https://tic.msu.edu/tgif/flink?recno=151038 |
| If there are problems with this record, send us feedback about record 151038. |
| Choices for finding the above item: |
| Web URL(s): https://archive.lib.msu.edu/tic/its/articles/2009jou437.pdf Last checked: 10/18/2011 Requires: PDF Reader |
| MSU catalog number: b2548899 |
| Request through your local library's inter-library loan service (bring or send a copy of this TGIF record) |