Abstract
This paper presents digital image preprocessing and handwritten digits recognition integrated in single FPGA chip. The application has been designed under Handel-C (DK1.0 environment). Presented solution integrates 3 classical stages of image processing: pre-processing, medium level analysis (Le. segmentation, labeling), pattern recognition and classification performed by an artificial neural network. Applied algorithms have different complexity levels, consume different types and amounts of FPGA resources. Special attention has been focused on coding and integration methodology of different algorithms in reprogrammable structure. Several problems related to implementation and optimization process have been described in the paper.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
