DocumentCode
1945511
Title
Reading Recognition of Digital Display Instrument Based on BP Neutral Network
Author
Huang, Zhiqiang ; Wan, Chengfu
Author_Institution
Sch. of Comput. Sci. & Technol., North China Electr. Power Univ., Baoding
Volume
1
fYear
2008
fDate
12-14 Dec. 2008
Firstpage
106
Lastpage
109
Abstract
To recognize digital display instrumentpsilas reading, a BP neutral network is designed, an improved BP algorithm and 15-feature extraction method is proposed. The image of instrument plate is obtained by an digital camera and transmitted to PC firstly, then an image preprocessing is carried through. The image preprocessing includes gray processing, binarization, grads sharp, vertical tilt correction, noise point removal, character segmentation, character unitary adjustment, character contraction and feature extraction. Character features are extracted by 15-feature extraction method. Finally recognize the reading by the BP neutral network. The experiment result proves that the system can scan and recognize lots of instruments, the recognition is of high veracity and great efficienc.
Keywords
backpropagation; cameras; computerised instrumentation; digital instrumentation; display instrumentation; feature extraction; image denoising; image segmentation; neural nets; backpropagation neural network algorithm; character contraction; character segmentation; character unitary adjustment; digital camera; digital display instrument reading recognition; feature extraction; grads sharp; gray processing; image preprocessing; noise point removal; vertical tilt correction; Computer displays; Computer science; Digital cameras; Digital control; Feature extraction; Image recognition; Image segmentation; Instruments; Lenses; Monitoring; BP neutral network; digital display instrument; reading recognition;
fLanguage
English
Publisher
ieee
Conference_Titel
Computer Science and Software Engineering, 2008 International Conference on
Conference_Location
Wuhan, Hubei
Print_ISBN
978-0-7695-3336-0
Type
conf
DOI
10.1109/CSSE.2008.778
Filename
4721703
Link To Document