Baseline drift removal method and device and median filter

Abstract

The invention relates to a baseline drift removal method and device and a median filter. The baseline drift removal method comprises the following steps: acquiring sampling point signal values with preset number to form an initial sort column; sorting the initial sort column from small to large to obtain an initial median; acquiring the next sampling point signal value, putting at the end of the initial sort column, and deleting a first sampling point signal value in the initial sort column to form a current sort column; acquiring the median of the current sort column according to the next sampling point signal value and the initial median; and calculating the signal value of removed baseline drift according to the median of the current sort column and the next sampling point signal value. In the baseline drift removal method and device and the median filter, all signal sampling points in each filter window do not need to be sorted, thereby, the baseline drift removal process is simply, conveniently and rapidly accelerated; and when the median is acquired, the next sampling point signal value and the median in the last sort column are selectively judged in sequence of magnitude, and further, the speed of a fast algorithm is increased.

Description

technical field [0001] The present invention relates to the technical field of signal baseline drift, in particular to a median filter-based baseline drift removal method, device and median filter. Background technique [0002] The phenomenon of signal baseline drift widely exists in many fields such as communication and medical treatment. Drift is a slow upward or downward movement of the baseline, which can be observed for a long time and can often mask noise and small fluctuations, seriously affecting the Subsequent processing of the signal. [0003] Taking TV signals as an example, in the process of collecting, amplifying, detecting and recording TV signals, there are often various interferences from the outside world. Rapid changes, superimposing a kind of upward or downward slow movement, because the long-term superimposition is very large, it can often mask noise and small fluctuations, and even make the output signal exceed the value range of the signal value, which...

AI Technical Summary

Problems solved by technology

[0005] However, when using the median filtering method directly, in the case of a long filtering window (such as when the filter length N=4096 signal sampling points), due to the need to calculate the median, it is first necessary to calculate the median in each filter window 4096 signal sampling points are sorted. The sorting algorithm runs slowly. Even if a faster quick sorting algorithm is used, it is often difficult to meet the requirements of the system design.

Images