Databases Reference
In-Depth Information
FIGURE 8.6: The Karajan grid engine architecture.
racy of the DHGN multi-feature scheme applied to face recognition have been
presented in [92].
8.2.1 Color and Spatio-Structural Features Consideration
The DHGN multi-feature scheme takes a holistic approach to incorporating
simultaneously the color and spatio-structural features into the recognition
process. A binary signature scheme proposed by Nascimento and Chitkara
[93] was adopted for content-based image retrieval (CBIR) in the pattern
recognition procedure. This scheme integrates the global binary signature with
an edge detection technique, such as Sobel's [94], for DHGN single-cycle image
recognition. In this multi-feature recognition approach, any number of features
can be included for pattern analysis by incorporating a su
cient number of
DHGN networks.
8.2.1.1
Global Binary Signature Scheme for Color Recognition
A common approach used to represent color distributions of an image is the
Global Color Histogram (GCH). Given an n-color model, a GCH is developed
