adaptive threshold
in [Matlab]
|
From: alkan alkaya on 19 Dec 2009 16:29 How can I apply an adaptive threshold to signal (one dimension signal).
From: ImageAnalyst on 19 Dec 2009 17:11 On Dec 19, 4:29 pm, "alkan alkaya" <alkanalk...(a)hotmail.com> wrote: > How can I apply an adaptive threshold to signal (one dimension signal). ----------------------------------------------------------------------------------------------------------- You can use a for loop to change the threshold as a function of index number in the array. The threshold may vary depending on only the index number (probably not likely), or it may vary depending on what is in the signal in a window around the current index (the more common situation).
From: alkan alkaya on 19 Dec 2009 17:29 can you explain me clearly or give me an example please. thanks..
From: ImageAnalyst on 19 Dec 2009 17:55 % Demo for one arbitrary way to adaptively threshold a signal % by ImageAnalyst % IMPORTANT: Be sure to join any lines broken into two by the newsreader. clc; close all; clear all; % Create sample data in the range 20-100 signalLength = 50; signal = 100*rand(signalLength, 1) + 20; subplot(2, 1, 1); bar(signal); % Set x axis to end at last bar. set(gca,'xlim',[1 signalLength]) title('Original signal'); set(gcf, 'Position', get(0,'Screensize')) ; % Maximize figure window. thresholdedSignal = signal; windowWidth = 7; halfWidth = floor(windowWidth/2); % Do an adaptive threshold. % As an example, this gives an output equal to the input % if the signal is above the mean in a window. % If the signal is below the mean it sets the output equal to 10; % Totally arbitrary - you have to decide what you want to do. thresholdValues = zeros(1, signalLength); for index = (halfWidth + 1) : (signalLength - halfWidth) leftIndex = index - halfWidth; rightIndex = index + halfWidth; signalWithinWindow = signal(leftIndex : rightIndex); meanInWindow = mean(signalWithinWindow); thresholdValues(index) = meanInWindow; if signal(index) > meanInWindow thresholdedSignal(index) = signal(index); else thresholdedSignal(index) = 10; end end subplot(2, 1, 2); bar(thresholdedSignal); % Set x axis to end at last bar. set(gca,'xlim',[1 signalLength]) hold on; plot(thresholdValues, 'ro'); title('Example of one way to adaptively threshold a signal');
From: alkan alkaya on 20 Dec 2009 17:07
thank you very much ImageAnalyst. I will try to use this program for my input signal. Have you got any other way to adaptively threshold a signal? and what is this methods name? again thank you very very much. ImageAnalyst <imageanalyst(a)mailinator.com> wrote in message <c6fe1e95-489d-48d3-9fef-107228ee1362(a)j24g2000yqa.googlegroups.com>... > % Demo for one arbitrary way to adaptively threshold a signal > % by ImageAnalyst > % IMPORTANT: Be sure to join any lines broken into two by the > newsreader. > clc; > close all; > clear all; > % Create sample data in the range 20-100 > signalLength = 50; > signal = 100*rand(signalLength, 1) + 20; > subplot(2, 1, 1); > bar(signal); > % Set x axis to end at last bar. > set(gca,'xlim',[1 signalLength]) > title('Original signal'); > set(gcf, 'Position', get(0,'Screensize')) ; % Maximize figure window. > > thresholdedSignal = signal; > windowWidth = 7; > halfWidth = floor(windowWidth/2); > % Do an adaptive threshold. > % As an example, this gives an output equal to the input > % if the signal is above the mean in a window. > % If the signal is below the mean it sets the output equal to 10; > % Totally arbitrary - you have to decide what you want to do. > thresholdValues = zeros(1, signalLength); > for index = (halfWidth + 1) : (signalLength - halfWidth) > leftIndex = index - halfWidth; > rightIndex = index + halfWidth; > signalWithinWindow = signal(leftIndex : rightIndex); > meanInWindow = mean(signalWithinWindow); > thresholdValues(index) = meanInWindow; > if signal(index) > meanInWindow > thresholdedSignal(index) = signal(index); > else > thresholdedSignal(index) = 10; > end > end > subplot(2, 1, 2); > bar(thresholdedSignal); > % Set x axis to end at last bar. > set(gca,'xlim',[1 signalLength]) > hold on; > plot(thresholdValues, 'ro'); > title('Example of one way to adaptively threshold a signal'); |