CVIPtools

The Computer Vision and Image Processing Feature Extraction and Pattern Classification Tool, CVIP-FEPC, is used to advance human and computer vision applications. While its primary function is computer vision, it serves various purposes, such as supporting the development of image compression schemes for human vision applications by identifying image features needed for a specific compression scheme. In computer vision applications, feature extraction and pattern classification form the fundamental components of image analysis systems for particular applications.

CVIP-FEPC's primary function is batch processing of large image sets. Users can select features and pattern classification parameters to process these image sets automatically. CVIP-FEPC allows users to define training and test sets and conduct multiple experiments to identify parameters for classifying objects of interest in the photos.

CVIP-FEPC is primarily designed to work with images containing binary masks representing individual objects of interest, with one object per image. It is compatible with manually created masks using CVIP tools or other tools available in various image database applications. Users load images, specify classes, select features and the test set, choose pattern classification parameters, and initiate the processing of the image set. Subsequently, an output file containing the experiment results is generated.^[3]

The CVIP Toolbox contains library functions for use in MATLAB. It includes arithmetic and logic functions, banding, coloring, conversion of image files, edge/line detection, geometry, histograms, mapping, morphological functions, noise filters, objective fidelity metrics, pattern classification and classification algorithms, segmentation, spatial filters, and transformation filters. The CVIP Toolbox also includes a CVIPlab skeleton program and a utility function for batch processing.^[4]