The Image-seg dataset
The information is a replica of the notes for the segmentation dataset
from the UCI
1. Title: Image Segmentation data
2. Source Information
- Creators: Vision Group, University of Massachusetts
- Donor: Vision Group (Carla Brodley, email@example.com)
- Date: November, 1990
3. Past Usage: None yet published
4. Relevant Information:
The instances were drawn randomly from a database of 7 outdoor
images. The images were handsegmented to create a classification
for every pixel.
Each instance is a 3x3 region.
5. Number of Instances: Training data: 210 Test data: 2100
6. Number of Attributes: 19 continuous attributes
7. Attribute Information:
- region-centroid-col: the column of the center pixel of the region.
- region-centroid-row: the row of the center pixel of the region.
- region-pixel-count: the number of pixels in a region = 9.
- short-line-density-5: the results of a line extraction algorithm that
counts how many lines of length 5 (any orientation) with
low contrast, less than or equal to 5, go through the region.
- short-line-density-2: same as short-line-density-5 but counts lines
of high contrast, greater than 5.
- vedge-mean: measure the contrast of horizontally
adjacent pixels in the region. There are 6, the mean and
standard deviation are given. This attribute is used as
a vertical edge detector.
- vegde-sd: (see 6)
- hedge-mean: measures the contrast of vertically adjacent
pixels. Used for horizontal line detection.
- hedge-sd: (see 8).
- intensity-mean: the average over the region of (R + G + B)/3
- rawred-mean: the average over the region of the R value.
- rawblue-mean: the average over the region of the B value.
- rawgreen-mean: the average over the region of the G value.
- exred-mean: measure the excess red: (2R - (G + B))
- exblue-mean: measure the excess blue: (2B - (G + R))
- exgreen-mean: measure the excess green: (2G - (R + B))
- 17. value-mean: 3-d nonlinear transformation
of RGB. (Algorithm can be found in Foley and VanDam, Fundamentals
of Interactive Computer Graphics)
- saturation-mean: (see 17)
- hue-mean: (see 17)
8. Missing Attribute Values: None
9. Class Distribution:
- Classes: brickface, sky, foliage, cement, window, path, grass.
- 30 instances per class for training data.
- 300 instances per class for test data.
10. Modifications for Delve
- The data and test files were combined and then
stratified to ensure equal representation of the output classes in each of the
Delve task-instance training sets.
- Attribute 3 (region-pixel-count) was deleted since it is a constant for