A New Method for Real-time Wavelet De-noising

This paper presents a new method for a real-time sequence with multi-scale wavelet transform analysis. When the amount of the sequence data is too small, using an artificial noise sequence with a certain variance and expend the sequence data to sufficient amount. Then the sequence real value can be recovered by off-line process and driving out some redundancy data. As for those data increasing with the time, supplementing and eliminating the redundancy data with a certain regularity to avoid the de-noising time with exceeding the renewal rate of the real-time system. This method is easy to be executed and suitable for digital process and the application of the computers.