1. TITLE:
 Vehicle silhouette dataset

 PURPOSE
 to classify a given silhouette as one of four types of vehicle,
 using a set of features extracted from the silhouette. The
 vehicle may be viewed from one of many different angles.


2. USE IN STATLOG

 2.1 Testing Mode
 9Fold Cross Validation

 2.2 Special Preprocessing
 No

 2.3 Test Results
 Success Rate TIME
 Algorithm Train Test Train Test
 
 QuaDisc 91.5 85.000 251 29
 Dipol92 ? 84.900 ? 1
 Alloc80 100 82.700 30 10
 LogDisc 83.3 80.800 758 8
 BackProp 83.2 79.300 14411 4
 Discrim 79.8 78.400 16 3
 Smart 93.8 78.300 3017 1
 Cart ? 76.500 29 1
 C4.5 93.5 73.400 153 1
 BayTree ? 72.900 4 1
 KNN 100 72.500 164 23
 Cal5 70.3 72.100 41 1
 Cascade ? 72.000 ? 1
 LVQ ? 71.300 ? ?
 Ac2 ? 70.400 595 23
 IndCart 95.3 70.200 85 1
 NewId 97 70.200 18 1
 Radial 90.2 69.300 1736 12
 Cn2 98.2 68.600 100 1
 Itrule ? 67.600 985 ?
 Kohonen 88.5 66.000 5962 50
 Castle 49.5 49.500 23 3
 Bayes 48.1 44.200 4 1
 Default ? 25.000


3. SOURCES and PAST USAGE

 ORIGINAL SOURCE
 Drs.Pete Mowforth and Barry Shepherd
 Turing Institute
 George House
 36 North Hanover St.
 Glasgow
 G1 2AD

 This dataset comes from the Turing Institute, Glasgow, Scotland.
 If you use this dataset in any publication please acknowledge this
 source.

 HISTORY
 This data was originally gathered at the TI in 198687 by
 JP Siebert. It was partially financed by Barr and Stroud Ltd.
 The original purpose was to find a method of distinguishing
 3D objects within a 2D image by application of an ensemble of
 shape feature extractors to the 2D silhouettes of the objects.
 Measures of shape features extracted from example silhouettes
 of objects to be discriminated were used to generate a class
 ification rule tree by means of computer induction.
 This object recognition strategy was successfully used to
 discriminate between silhouettes of model cars, vans and buses
 viewed from constrained elevation but all angles of rotation.
 The rule tree classification performance compared favourably
 to MDC (Minimum Distance Classifier) and kNN (kNearest Neigh
 bour) statistical classifiers in terms of both error rate and
 computational efficiency. An investigation of these rule trees
 generated by example indicated that the tree structure was
 heavily influenced by the orientation of the objects, and grouped
 similar object views into single decisions.

 DESCRIPTION
 The features were extracted from the silhouettes by the HIPS
 (Hierarchical Image Processing System) extension BINATTS, which
 extracts a combination of scale independent features utilising
 both classical moments based measures such as scaled variance,
 skewness and kurtosis about the major/minor axes and heuristic
 measures such as hollows, circularity, rectangularity and
 compactness.
 Four "Corgie" model vehicles were used for the experiment:
 a double decker bus, Cheverolet van, Saab 9000 and an Opel Manta 400.
 This particular combination of vehicles was chosen with the
 expectation that the bus, van and either one of the cars would
 be readily distinguishable, but it would be more difficult to
 distinguish between the cars.
 The images were acquired by a camera looking downwards at the
 model vehicle from a fixed angle of elevation (34.2 degrees
 to the horizontal). The vehicles were placed on a diffuse
 backlit surface (lightbox). The vehicles were painted matte black
 to minimise highlights. The images were captured using a CRS4000
 framestore connected to a vax 750. All images were captured with
 a spatial resolution of 128x128 pixels quantised to 64 greylevels.
 These images were thresholded to produce binary vehicle silhouettes,
 negated (to comply with the processing requirements of BINATTS) and
 thereafter subjected to shrinkexpandexpandshrink HIPS modules to
 remove "salt and pepper" image noise.
 The vehicles were rotated and their angle of orientation was measured
 using a radial graticule beneath the vehicle. 0 and 180 degrees
 corresponded to "head on" and "rear" views respectively while 90 and
 270 corresponded to profiles in opposite directions. Two sets of
 60 images, each set covering a full 360 degree rotation, were captured
 for each vehicle. The vehicle was rotated by a fixed angle between
 images. These datasets are known as e2 and e3 respectively.
 A further two sets of images, e4 and e5, were captured with the camera
 at elevations of 37.5 degs and 30.8 degs respectively. These sets
 also contain 60 images per vehicle apart from e4.van which contains
 only 46 owing to the difficulty of containing the van in the image
 at some orientations.

4. ATTRIBUTE DISCRIPTION

 NUMBER OF EXAMPLES

 Total no. = 846
 (946 in original dataset, 100 examples
 were kept by Strathclyde for validation.
 So the dataset issued contains only 846 examples).
 NUMBER OF CLASSES

 4 OPEL, SAAB, BUS, VAN

 Class Nr.Examples
 
 1 212 (25.06%)
 2 217 (25.65%)
 3 218 (25.77%)
 4 199 (23.52%)

 NUMBER OF ATTRIBUTES

 No. of atts. = 18

 COMPACTNESS (average perim)**2/area

 CIRCULARITY (average radius)**2/area

 DISTANCE CIRCULARITY area/(av.distance from border)**2

 RADIUS RATIO (max.radmin.rad)/av.radius

 PR.AXIS ASPECT RATIO (minor axis)/(major axis)

 MAX.LENGTH ASPECT RATIO (length perp. max length)/(max length)

 SCATTER RATIO (inertia about minor axis)/(inertia about major axis)

 ELONGATEDNESS area/(shrink width)**2

 PR.AXIS RECTANGULARITY area/(pr.axis length*pr.axis width)

 MAX.LENGTH RECTANGULARITY area/(max.length*length perp. to this)

 SCALED VARIANCE (2nd order moment about minor axis)/area
 ALONG MAJOR AXIS

 SCALED VARIANCE (2nd order moment about major axis)/area
 ALONG MINOR AXIS

 SCALED RADIUS OF GYRATION (mavar+mivar)/area

 SKEWNESS ABOUT (3rd order moment about major axis)/sigma_min**3
 MAJOR AXIS

 SKEWNESS ABOUT (3rd order moment about minor axis)/sigma_maj**3
 MINOR AXIS

 KURTOSIS ABOUT (4th order moment about major axis)/sigma_min**4
 MINOR AXIS

 KURTOSIS ABOUT (4th order moment about minor axis)/sigma_maj**4
 MAJOR AXIS

 HOLLOWS RATIO (area of hollows)/(area of bounding polygon)

 Where sigma_maj**2 is the variance along the major axis and
 sigma_min**2 is the variance along the minor axis, and

 area of hollows= area of bounding polyarea of object

 The area of the bounding polygon is found as a side result of
 the computation to find the maximum length. Each individual
 length computation yields a pair of calipers to the object
 orientated at every 5 degrees. The object is propagated into
 an image containing the union of these calipers to obtain an
 image of the bounding polygon.


BIBLIOGRAPHY

 Turing Institute Research Memorandum TIRM87018 "Vehicle
 Recognition Using Rule Based Methods" by Siebert,JP (March 1987)



CONTACTS
 statlogadm@ncc.up.pt
 bob@stams.strathclyde.ac.uk


================================================================================

1,2,3,4.
COMPACTNESS: continuous.
CIRCULARITY: continuous.
DISTANCE CIRCULARITY: continuous.
RADIUS RATIO: continuous.
PR AXIS ASPECT RATIO: continuous.
MAX LENGTH ASPECT RATIO: continuous.
SCATTER RATIO: continuous.
ELONGATEDNESS: continuous.
PR AXISRECTANGULAR: continuous.
LENGTHRECTANGULAR: continuous.
MAJORVARIANCE: continuous.
MINORVARIANCE: continuous.
GYRATIONRADIUS: continuous.
MAJORSKEWNESS: continuous.
MINORSKEWNESS: continuous.
MINORKURTOSIS: continuous.
MAJORKURTOSIS: continuous.
HOLLOWS RATIO: continuous.