Table 1.

The list of 27 features extracted from each root crown image by the RhizoVision Analyzer.

Features extracted	Description
Median and maximum number of roots	The number of roots is counted by performing horizontal line scans from left to right in each row through the segmented image. In each of the line scan, pixel value transitions are checked from the previous pixel value to the current pixel value on its right side. If the current pixel value changes from 0 to 1, a root is recorded. The number of roots is recorded from each row of the segmented image, and the median and maximum numbers of roots are determined from these values.
Number of root tips	Count of the total number of tip pixels in the skeletonized image.
Total root length	Computed by determining the total Euclidean distance of medial axis pixels in the skeletonized image.
Depth, maximum width, and width-to-depth ratio	The maximum depth and maximum width of the root in the segmented image. The ratio of maximum width to depth of the image is noted as the width-to-depth ratio.
Network area, convex area, and solidity	The total number of pixels in the segmented image. The convex hull of the segmented image is the minimal-sized convex polygon that can contain the root. From this convex hull, the convex area is determined. The ratio of the network area and the convex area is given as the solidity.
Perimeter	Perimeter is the total Euclidean distance of contour pixels in the segmented image.
Average, median, and maximum diameters	For each pixel in the skeletonized image, the distance to the nearest nonroot pixel is computed. This distance is used as a radius to fit a circle. The diameter of the circle at each pixel is noted as the root diameter. The list of diameters from all the medial axis pixels is used to determine the average, median, and maximum diameters for the entire root crown.
Volume and surface area	Using the radii determined earlier, the sum of all cross-sectional areas across all the medial axis pixels is noted as the volume and the sum of the perimeter across all the medial axis pixels is noted as the surface area.
Lower root area	The lower root area is the area of the segmented image pixels that are located below the location of the medial axis pixel that has the maximum radius.
Holes and average hole size	Holes are the disconnected background components and indicative of root branching and complexity. They can be counted (number) by inverting the segmented image. The average hole size (area) is also calculated.
Average root orientation	For every medial axis pixel, the orientation at the pixel is computed by determining the mean orientation of medial axis pixels in a 40 × 40 pixel locality. The average of all these orientations is noted as the average root orientation.
Fine, medium, and coarse diameter frequencies	From the skeletal image, the medial axis pixels are grouped into fine or coarse roots based on the diameter values at the pixels.
Shallow, medium, and steep angle frequencies	Given the skeletal image, for every pixel in the medial axis, the locations of the medial axis pixels in a 40 × 40 pixel locality are used to determine the orientation of these pixels in the locality. This orientation is noted for every medial axis pixel. These orientation frequencies are grouped in bins less than 30, less than 60, and less than 90 degrees from the horizontal.
Computational time	The time taken to extract features for one image.