透過您的圖書館登入
IP:18.117.158.47
  • 期刊
  • OpenAccess

Computer vision-based Object-recognition by a Hopfield Neural Network

運用霍普菲爾類神經網路之電腦視覺物件識別

並列摘要


In this study, a computer vision-based object-recognition system using a Hopfield neural network and Visual Basic language is proposed. The system function for object recognition, invariant under translation and scaling, is addressed also. As system parameters such as background image, template image, and test image can be selected quickly with operation in Windows 98/NT platforms, computations for learning and recalling of the Hopfield neural network will be further finished in a couple of seconds. The proposed study is the primary stage of object-recognition camera system capable to be extended for various real-time applications such as infrared missile seeker, infrared coast security, industrial quality control, and home security.

延伸閱讀