Hi, I doing a project that load a roadsigns.mat, which a exported from Image Lableler. Can someone help me to fix the error below ?

'Error using trainCascadeObjectDetector>parseInputs (line 310) The value of 'positiveInstances' is invalid. Expected input number 2, positiveInstances, to be one of these types: struct, char'

load('roadSignals.mat');
  imDir = fullfile('/Users/Anisio/Documents/MATLAB/IMAGE');

    %Add the images location to the MATLAB path.
    addpath(imDir);

%Specify the folder for negative images.
negativeFolder = fullfile('/Users/Anisio/Documents/MATLAB/IMAGE');

%Train a cascade object detector called 'stopSignDetector.xml' using HOG features. The following command may take several minutes to run:
trainCascadeObjectDetector('SignDetector.xml',labelingSession,negativeFolder,'FalseAlarmRate',0.2,'NumCascadeStages',5);

%Use the newly trained classifier to detect a stop sign in an image.
detector = vision.CascadeObjectDetector('stopSignDetector.xml');

%Read the test image.
img = imread('SpeedCameraSample.jpg');

%Detect a stop sign.
bbox = step(detector,img);

%Insert bounding boxes and return marked image.
detectedImg = insertObjectAnnotation(img,'rectangle',bbox,'stop sign');

%Display the detected stop sign.
figure;
imshow(detectedImg);

%Remove the image directory from the path.
rmpath(imDir);

Answer 1

[Dima Lisin]

You should export the ROIs using the "Export" button. That will create a variable in the workspace that you can pass into trainCascadeObjectDetector.

You have saved a labeling session. It is for resuming labeling after a break, not for passing into a function.

[yang liu]

I had the same problem. As you described above, I then use "Export ROIs", and gave a variable named as "data".

It gives me the following error: "Error using trainCascadeObjectDetector (line 245) Unable to determine file format."

Attached image shows all the cmd, error, and all.

[Dima Lisin]

It looks like one of your negative images may be corrupt. Please check what's in your nonHardHat directory, and make sure you can read all of the images from it into MATLAB with imread.