what means this error , while trying to use vision.blobAnalysis to detect motions?

what means this error , while trying to use vision.blobAnalysis to detect motions?

Error using coder.internal.errorIf (line 8)
The POSITION matrix must have four columns for shape Rectangle

Error in insertShape>errIf1 (line 796)
coder.internal.errorIf(condition, msgID, strArg);

Error in insertShape>crossCheckShapePosition (line 347)
     errIf1(errCond, 'vision:insertShape:posColsNot4ForRect', shape);

Error in insertShape>crossCheckInputs (line 317)
crossCheckShapePosition(shape, position);

Error in insertShape>validateAndParseInputs (line 176)
    crossCheckInputs(shape2, position, color);

Error in insertShape (line 103)
    validateAndParseInputs(I, shape, position, varargin{:});

Error in btw (line 134)
result = insertShape(frame, 'Rectangle', bbox, 'Color', 'green');

Answer 1

[Walter Roberson]

Check size(bbox) . The bbox that you are passing in must have 4 columns, such as a row vector of 4 elements.

[syhem samti]

blob generation works well in my code, but when i try to extract MajorAxisLength, MinorAxisLength of each blob...this error appears

[Walter Roberson]

Please show your call to vision.BlobAnalysis and your step() command

[Image Analyst]

This is not what you originally said. You said there was a problem with rectangle and bbox. Did you solve that problem and now have this new one???? Aren't MajorAxisLength and MinorAxisLength computed by regionprops()? You don't even show that code or even mention that in your original question. It's hard for people to answer questions about code that is not shown, don't you think?

[Walter Roberson]

Image Analyst: see http://www.mathworks.com/help/vision/ref/vision.blobanalysis-class.html

The blob analysis can be configured to output major and minor axis lengths, but the order of the outputs becomes positional according to the exact options you have enabled, rather than by structure. The order of the positional outputs is not well defined in the blobanalysis step() documentation, and would have to be presumed that any one further down the list would appear to the right of any one further up higher in the list.

Chances are that the poster is expecting the outputs in the wrong order and passing what they think should be the bbox output but is a different output.

[Image Analyst]

I don't have that toolbox. I know the Computer Vision Toolbox requires the Image Processing Toolbox so I wouldn't be surprised if it called regionprops() internally.

The link you gave says "The order of the returned values when there are multiple outputs are in the order they are described below:"

It looks like if you call

[AREA,CENTROID,BBOX] = step(H,BW)

you get those 3 results. Then, if you want any others, you have to call them in a very specific order - the order listed in the help. It seems weird and inconvenient that you have to call

[___, someResult] = step(H,BW);

a bunch of times serially to get what you want rather than just asking for it directly like you can with regionprops(). Anyway, the next two after those 3 are MajorAxis and MinorAxis.

Again, the poster does not seem to show any of this code where MajorAxis is asked for.

[Walter Roberson]

You do not call serially, it is

[AREA*, CENTROID*, BBOX*, MAJORAXIS*, MINORAXIS*, ORIENTATION*...] = step(H, BW)

where each * is intended to indicate here that the corresponding output might be omitted depending on which *OutputPort you configured. For example you might end up with

[CENTROID, MAJORAXIS] = step(H,BW)

if you turned off AreaOutputPort and BoundingBoxOutputPort and turned on MajorAxisLengthOutputPort .

Vision.BlobAnalysis works in the System Objects paradigm, with the ability to generate optimized code that reduces memory transfers. I doubt it calls into regionprops().

[syhem samti]

here the code is:

      function videoAnalysis()
  clear all;
  close all;
  clc;
  foregroundDetector = vision.ForegroundDetector('NumGaussians', 3, 'NumTrainingFrames', 50);

videoReader = vision.VideoFileReader('video6.avi');
% fRate = videoReader.info.VideoFrameRate;
% disp(fRate);

for i = 1:50
    frame = step(videoReader);
    foreground = step(foregroundDetector, frame);

end
imwrite(foreground,'foreground1.jpg')
BW = imread('foreground1.jpg');
%Calculate the perimeters of objects in the image.

BWs = bwperim(BW,8);
%Display the original image and the perimeters side-by-side.

%imshowpair(BW,BWs,'montage')

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
se90 = strel('line', 3, 90);
se0 = strel('line', 3, 0);

%The binary gradient mask is dilated using the vertical structuring element followed by the horizontal structuring element. The imdilate function dilates the image.

BWsdil = imdilate(BWs, [se90 se0]);
%figure, imshow(BWsdil), title('dilated gradient mask');

BWdfill = imfill(BWsdil, 'holes');
% figure, imshow(BWdfill);
% title('binary image with filled holes');

BWnobord = imclearborder(BWdfill, 4);
%figure, imshow(BWnobord), title('cleared border image');

seD = strel('diamond',9);
BWfinal = imerode(BWnobord,seD);
BWfinal = imerode(BWfinal,seD);
%figure, imshow(BWfinal), title('segmented image');

BWoutline = bwperim(BWfinal);
Segout = BW;
Segout(BWoutline) = 255;
%figure, imshow(Segout), title('outlined original image');

se = strel('square', 3);
fF = imopen(Segout, se)

se1 = strel('square', 5);%masque
ForF = imdilate(fF, se1);
%figure; imshow(ForF); title('finaal foreground ')

L = bwlabel(ForF);% connected component (seulement sur les images binaires
s = regionprops(L, 'Area');
s(1)
area_values = [s.Area]
idx3 = find((5500 <= area_values) & (area_values <= 1900000))
bw3 = ismember(L, idx3);
figure, imshow(bw3);title(' finaal foreground ')

se = strel('square', 3);
filteredForeground = imdilate(bw3, se)
figure, imshow(filteredForeground); 
title('Fforeground')

blobAnalysis = vision.BlobAnalysis('BoundingBoxOutputPort', true, ...
    'AreaOutputPort', false, 'CentroidOutputPort', false, ...
    'MinimumBlobArea', 350 , 'MaximumBlobArea', 15000, ...
    'MajorAxisLengthOutputPort' , true, 'MaximumCount', 5);

[centroid, bbox] = step(blobAnalysis, filteredForeground)
result = insertShape(frame, 'Rectangle', bbox, 'Color', 'green');

numPerso = size(bbox, 1);
result = insertText(result, [10 10], numPerso, 'BoxOpacity', 1, 'FontSize', 14);

videoPlayer = vision.VideoPlayer('Name', 'Detected Cars');

videoPlayer.Position(3:4) = [650,400];  % window size: [width, height]
% se = strel('square', 3); % morphological filter for noise removal

i = 0;

while ~isDone(videoReader)
%     if ( mode(i,10) == 0 )
%         disp(i);
%     end
%     i = i + 1;
    frame = step(videoReader); % read the next video frame

      % Detect the foreground in the current video frame
      foreground = step(foregroundDetector, frame);

      % Use morphological opening to remove noise in the foreground
  %     filteredForeground = imopen(foreground, se);
  %%by me

se = strel('square', 3);
filteredForeground = imdilate(bw3, se)
%figure, imshow(filteredForeground); 
%title('Fforeground')
      % Detect the connected components with the specified minimum area, and
      % compute their bounding boxes

      % Draw bounding boxes around the detected cars
      result = insertShape(frame, 'Rectangle', bbox, 'Color', 'blue');
  %     imshow(result);
  %     

      % Display the number of cars found in the video frame
      numCars = size(bbox, 1);
      result = insertText(result, [10 10], numPerso, 'BoxOpacity', 1, 'FontSize', 14);

      %if there is moving object
      if numPerso == 0
          result = insertText(result, [100 20], 'non motion detected', 'BoxOpacity', 1, 'FontSize', 14, 'BoxColor', 'green');
      else

          result = insertText(result, [100 20], 'motion detected', 'BoxOpacity', 1, 'FontSize', 14, 'BoxColor', 'red');
      end

      step(videoPlayer, result);  % display
  end

release(videoReader); % close
  end

[Walter Roberson]

Look at your code again:

blobAnalysis = vision.BlobAnalysis('BoundingBoxOutputPort', true, ...
    'AreaOutputPort', false, 'CentroidOutputPort', false, ...
    'MinimumBlobArea', 350 , 'MaximumBlobArea', 15000, ...
    'MajorAxisLengthOutputPort' , true, 'MaximumCount', 5);

[centroid, bbox] = step(blobAnalysis, filteredForeground)

You have 'CentroidOutputPort', false so the centroid output is not going to be produced. The bounding box is going to be the first output (and there is not going to be a second output)

bbox = step(blobAnalysis, filteredForeground);

[syhem samti]

AND HOW can i extract MajorAxisLength PLEASE?

[Walter Roberson]

[bbox, MajorAxisLength] = step(blobAnalysis, filteredForeground);

I overlooked that you had that turned on as well.

[syhem samti]

thank you for you answers...i tried this too but another error appears...anyway i just found another way to extract those attributes

[NIHARIKA MITTAL]

Can you please share the other way. I am also stuck in this same situation