The Recognition of Liquid Drop Fingerprint Based on Immune Neural Network

The Liquid Drop Fingerprint (LDF) is a way to effectively reflect liquid characteristics by using feature of graph. The waveform analysis method is chose for characteristic extraction which regards the LDF as a wave including the peak and trough. To speed up the convergence and reduce the over-fitting, a population control mechanism, inspired by the natural immune system, has been constructed to analyze and identify the liquid. The new algorithm is presented, along with the immunological metaphors used as inspiration. The result shows that the new method of pattern recognition can improve the speed of the network and the detecting velocity efficiently.