showing image frames and statistics in figures

I have a "real time" application that generates a stream of images, which I show as a four-part subplot (a few times every second). I also want to show a text box that has some statistics. I have two issues.

1. The imshow for each subplot (and title) is to slow. I suspect too much re-layout. Since a single imshow outside of the subplot would keep up with 25 fps.
2. I am trying to show the text statistics as a figure. Probably stupid, but it does not work as subplot. Anyway switching between figures causes new windows to appear.

Yes, all of this is a Class object, ticker an event handler that is called every time a new image frame is ready.

              this.scope = figure();
              this.stats = figure();

    %%Private Methods
    methods
        function showScope(this)
            % show raw input frame
            figure(this.scope);
            tTotAvg = this.stepper.tTotalAvg;
            tMed = this.stepper.tMedian;
            tConv = this.stepper.tConvolve;
            tThresh = this.stepper.tThreshold;
            subplot(1,4,1);
            imshow(this.raw);
            stext = sprintf('Total GPU: %0.2f ms', tTotAvg);
            title(stext, 'Fontweight','Bold');
            subplot(1,4,2);
            imshow(this.background);
            stext = sprintf('Median Filter: %0.2f ms', tMed);
            title(stext, 'Fontweight','Bold');
            subplot(1,4,3);
            imshow(this.foreground, [50,255]);
            stext = sprintf('Convolution Filter: %0.2f ms', tConv);
            title(stext, 'Fontweight','Bold');
            subplot(1,4,4);
            imshow(this.mask, [0,1]);
            stext = sprintf('Thresholding: %0.2f ms', tThresh);
            title(stext, 'Fontweight','Bold');
        end
        function showStats(this)
            tTotAvg = this.stepper.tTotalAvg;
            tMed = this.stepper.tMedian;
            tConv = this.stepper.tConvolve;
            tThresh = this.stepper.tThreshold;
            tTrans = this.stepper.tTransfer;
            figure(this.stats);
            clf(this.stats);
            %imshow(ones(1,1));
            s = cell(1,5);
            s{1} = sprintf('Avg Total:  %0.2f\n', tTotAvg);
            s{2} = sprintf('Median:     %0.2f\n', tMed);
            s{3} = sprintf('Convolve:   %0.2f\n', tConv);
            s{4} = sprintf('Threshold:  %0.2f\n', tThresh);
            s{5} = sprintf('Transfer:   %0.2f\n', tTrans);
            stext = sprintf('%s%s%s%s%s',s{1},s{2},s{3},s{4},s{5});
            text(0,0,stext);
        end
        function ticker(this, src, e)
            % Event Handler for each frame processed by the SkyMarker
            bits = uint8(this.stepper.raw);
            this.raw = reshape(bits, [this.height,this.width]);
            bits = uint8(this.stepper.background);
            this.background = reshape(bits, [this.height,this.width]);
            bits = single(this.stepper.foreground);
            this.foreground = reshape(bits, [this.height,this.width]);
            bits = uint8(this.stepper.mask);
            this.mask = reshape(bits, [this.height,this.width]);
            this.showScope();
            %this.showStats();
        end

[Ced]

When you need to update a plot a lot, the way to go is usually to save the plot handle, and then update the values by using "set", so something like this:

figure
h = plot([1 2], [1 2]);
set(h,'XData',[3 5])
drawnow

Not sure how to update this with full images though, but maybe this works too. 25 fps is a bit tricky... matlab may not be the best choice for that. But maybe somebody else has a better idea here.

Concerning the switching plots, there are two ways I can think of to fix that:

a) Give back the handle to the plot, and then always plot to the respective handle. This also works for subplots!

b) You can explicitely access a current figure through its number (e.g. call figure(2) to switch to that figure). Since version 2014a (or b, not sure), the number of a current figure (if you don't want to hardcode it) is a property and can be retrieved from the handle with get(handle,'Number').

Hint: if you have a handle, you can get a list of all possible settings by calling

set(handle)

without second argument

Hope this helps