Feature Vector Commands
   Subject Index Alphabetical Index General Information Home

The "vfeature" module is comprised of commands to facilitate the management and analysis of feature vector data. There is a simple k-nearest neighbor classifier in VisionX (vclasf). For other classifiers the VisionX commnads facilitate the generation and management of feature vectors in native format and their export in text format for use with other classifiers.

Feature vector data are stored in a VisionX file with the structure listed below. The commands can be separated into several broad categories: import/export to other systems and classifiers, feature vector file manipulation, and classification. The converion program vxltovf converts between teh VisionX binary feature vector format and comma separated value text files; it facilitates the linking between VisionX and other external classifiers.

Command Import/Export
vxltovf Converts a comma separated or tab separated file to VisionX feature vector format.
vxltovf -i Converts VisionX feature vector format file to a comma separated or tab separated file.
  Feature manipulation
vfvcat Concatenates feature vectors together
vfvpick Selects a subset of features from a feature file
vclasf Performs classification of feature vectors using a nearest neighbors algorithm

File Structure

All the programs utilize the VisionX feature vector file format. Feature vectors are stored in framed VisionX files according to the following structure (for each frame):

ElementData typeRequired?

A header is only allowed once per frame; it consists of a text lsit of all the feature names. Only one ID, class, and description are allowed for each feature vector; the programs assume that the ID and class will be present immediately preceding the feature vector in the specified order, the description is optional and is a free form description with only one constraint -- it can not contain "tab" or "," characters. Within a frame, all the feature vectors must have the same length. i

Output files from the standard nearest neighbor classification program (vclasf) have the following structure (for each frame):

ElementData typePresent?Comments
IDVX_IDonly if present in input
ClassVX_GSINTonly if present in input
DescriptionVX_DESConly if present in input
Class vectorVX_GINTalwaysFirst element is classified class, remaining elements are class of each of the nearest neighbors
Distance vectorVX_DISTalwaysFirst element is distance to classified class, remaining elements are distances to each of the nearest neighbors
ID vectorVX_IDVonly if ID present in input First element is ID of closest feature vector of classified class, remaining elements are ID of each of the nearest neighbors

 © 2019 Anthony P. Reeves