Title page for ETD etd-03022006-093418
|Type of Document
||Development of root observation method by image analysis system
||Biological Systems Engineering
|Vaughn, Darrel H.
|Deck, Sidney H.
|Kline, D. Earl
|Kulasiri, G. D.
|Perumpral, John V.
|Thomson, Steve J.
- image processing
- root observation method
|Date of Defense
Knowledge of plant roots is important for determining plant-soil relationships,
managing soil effectively, studying nutrient and water extraction, and creating a soil
quality index. Plant root research is limited by the large amount of time and labor
required to wash the roots from the soil and measure the viable roots. A root
measurement method based on image analysis was proposed to reduce the time and
labor requirement. A thinning algorithm-based image analysis method was used to
measure corn root length at the planar faces cut from a core sample. The roots were
exposed by careful handling and contrasted from the soil by causing autofluorescence
using long-wave ultraviolet light. The contrast-enhanced images were stored on the
camcorder video tape and digitized by frame grabber. A binary root image was
acquired from the digitized gray scale image by a thresholding operation. The binary
root image was thinned until the roots were reduced to their basic structure. Root
length was calculated from the number of pixels of the root's basic structure. This root
length was divided by the removed soil volume of the profile of the core sample to
estimate the root length density (RLD, cm root cm-3 soil). This estimated RLD was
regressed on RLD, measured from washed roots in the same soil core sample, and a
linear relationship (R2 = 0.96) was obtained. This study indicated that the image
analysis root measurement method can determine the length of corn root systems up to
2.5 times faster than by using the conventional method which incorporates a root
|| Approximate Download Time
| 28.8 Modem
|| 56K Modem
|| ISDN (64 Kb)
|| ISDN (128 Kb)
|| Higher-speed Access
next to an author's name indicates that all
files or directories associated with their ETD
are accessible from the Virginia Tech campus network only.
If you have questions or technical
problems, please Contact DLA.