Study On The Technology Of Digital Recognition In Medical Instrument

With the development and popularization of science and technology, the method on official management has gone to automatic or semi-automatic. In order to improve efficiency and reduce the artificial incorrect ruling, the technology of ICU monitor screen is studied, including real-time collection, identify and recording the related informational.Firstly, current research situation and application of Character Recognition are introduced. And the key techniques including image preprocessing, character segmentation and character recognition are discussed.By analyzing the feature of sampling image, the segmentation algorithm on color image by color clustering is used. The experimental results show the algorithm is effective on abstracting the foreground of character, and it is sensitive to the illumination conditions and irregular background. It has built a good foundation for character segmentation.Aiming at the problem of touching character, a segmentation method based on the bidirectional shortest-path is proposed by analyzing the behavior of touching character. Experimental results show it is effective.According to the Append Moment Term and Adaptive Adjustment the rate of learning, the traditional BP neural network is remolded for character recognition. Experiment shows that the method achieves a higher rate of correct identification.