Biomedical Engineering Reference
In-Depth Information
Fig. 3.1
The framework of
the proposed segmentation
framework
Input image
Preprocessing
Main processing algorithm
Quality Assurance process
End
3.3 Pre-processing
Before adopting the proposed Adaptive Crossed Reconstruction (ACR) segmenta-
tion algorithm, the input image has to undergo two pre-processing procedures to
serve two purposes: (1) to equalize the brightness of the image (2) to decrease the
variations in brightness different in bone structure. The first purpose is realized
by using a proposed histogram equalization specially designed for hand bone pre-
processing namely the Multipurpose Beta Optimized Bi-Histogram Equalization
Search WWH ::
Custom Search