Error: Undefined function 'preprocessData' for input arguments of type 'cell'.

I am implementing MATLAB 2019b examle "Object Detection Using YOLO v2 Deep Learning"(https://www.mathworks.com/help/vision/ug/train-an-object-detector-using-you-only-look-once.html), but when I run the following line of code:

anchorBoxes = estimateAnchorBoxes(preprocessedTrainingData ,numAnchors)

it gives me error as mentioned in title. Thanks for helping out.

A small piece of code from example is mentioned below:

%...............Code..................................%

unzip vehicleDatasetImages.zip

data = load('vehicleDatasetGroundTruth.mat');

vehicleDataset = data.vehicleDataset;

rng(0)

shuffledIdx = randperm(height(vehicleDataset));

idx = floor(0.6 * height(vehicleDataset));

trainingDataTbl = vehicleDataset(shuffledIdx(1:idx),:);

imdsTrain = imageDatastore(trainingDataTbl{:,'imageFilename'});

bldsTrain = boxLabelDatastore(trainingDataTbl(:,'vehicle'));

trainingData = combine(imdsTrain,bldsTrain);

inputSize = [224 224 3];

preprocessedTrainingData = transform(trainingData, @(data)preprocessData(data,inputSize));

numAnchors = 5;

anchorBoxes = estimateAnchorBoxes(preprocessedTrainingData ,numAnchors)

%.................................................................................................................................................%

[Harshveer Singh]

I am also getting the similar error. Unable to resolve it. I copied all the supporting function into a .m file. I then tried calling the function through its name in the command window but unable to do so. Someone please help how to go ahead with it.

Answer 1

[Steven Lord]

The preprocessData function is a supporting function that is part of that example and is not on the MATLAB path in general. Scroll down to the Supporting Functions section of that example and make a copy of that function, either as a local function inside the file where you're implementing your own variant of the example or as a separate function file.

[ali raza]

Thanks Steven, I have already tried it but still same. I separately save this local function in my working directly but issue didnt resolve yet.

[ali raza]

As per your suggestion, I embedded this function in my .m file, now its working fine. Thanks for your quick and fruitful response. :)

[TEJAS PHUTANE]

I have changed the code for detection of balls from 1000 images but still stuck on following problem while executing code:

trainingDataForEstimation = transform(trainingData,@(data)preprocessData(data,inputSize));

numAnchors = 7;

[anchorBoxes, meanIoU] = estimateAnchorBoxes(trainingDataForEstimation, numAnchors)

OUTPUT:

Invalid transform function defined on datastore.

The cause of the error was:

Error using bboxresize>iParseInputs (line 89)

The value of 'bboxA' is invalid. Expected input number 1, bboxA, to be integer-valued.

Error in bboxresize (line 49)

params = iParseInputs(bboxA,scale);

Error in training>preprocessData (line 56)

data{2} = bboxresize(data{2},scale);

Error in training>@(data)preprocessData(data,inputSize) (line 17)

trainingDataForEstimation = transform(trainingData,@(data)preprocessData(data,inputSize));

Error in matlab.io.datastore.TransformedDatastore/read (line 148)

data = self.Transforms{ii}(data);

Error in matlab.io.Datastore/readall (line 250)

data = [data; read(copyds)]; %#ok<AGROW>

Error in matlab.io.datastore.TransformedDatastore/readall (line 188)

data = [e-mail](copyds);

Error in estimateAnchorBoxes>iReadallBoxes (line 270)

boxes = readall(ds);

Error in estimateAnchorBoxes>iCheckBoxesFromDatastore (line 215)

boxes = iReadallBoxes(ds);

Error in estimateAnchorBoxes>iParseInputs (line 168)

boxes = iCheckBoxesFromDatastore(datastore);

Error in estimateAnchorBoxes (line 136)

[boxes, numAnchors, params] = iParseInputs(datastore, numAnchors, varargin{:});

Error in training (line 19)

[anchorBoxes, meanIoU] = estimateAnchorBoxes(trainingDataForEstimation, numAnchors)

[jose cabrera]

Hi, Did you find the solution??

Im getting the same problem

Answer 2

[michael scheinfeild]

some fix needed in preprocessing , first round and be sure it is positive top left , i think maybe check box outside image will be good , i didnt did it . also rounding of box cordinates is important !. also see that in image we have several bounding boxes and we check them all !

function data = preprocessData(data,targetSize)
% Resize image and bounding boxes to the targetSize.
scale = targetSize(1:2)./size(data{1},[1 2]);
data{1} = imresize(data{1},targetSize(1:2));

boxEstimate=round(data{2});
boxEstimate(:,1)=max(boxEstimate(:,1),1);
boxEstimate(:,2)=max(boxEstimate(:,2),1);

data{2} = bboxresize(boxEstimate,scale);
end

[michael scheinfeild]

same fo the next function

function B = augmentData(A)
% Apply random horizontal flipping, and random X/Y scaling. Boxes that get
% scaled outside the bounds are clipped if the overlap is above 0.25. Also,
% jitter image color.
B = cell(size(A));

I = A{1};
sz = size(I);
if numel(sz)==3 && sz(3) == 3
    I = jitterColorHSV(I,...
        'Contrast',0.2,...
        'Hue',0,...
        'Saturation',0.1,...
        'Brightness',0.2);
end

% Randomly flip and scale image.
tform = randomAffine2d('XReflection',true,'Scale',[1 1.1]);
rout = affineOutputView(sz,tform,'BoundsStyle','CenterOutput');
B{1} = imwarp(I,tform,'OutputView',rout);

% Apply same transform to boxes.

boxEstimate=round(A{2});
boxEstimate(:,1)=max(boxEstimate(:,1),1);
boxEstimate(:,2)=max(boxEstimate(:,2),1);


[B{2},indices] = bboxwarp(boxEstimate,tform,rout,'OverlapThreshold',0.25);
B{3} = A{3}(indices);

% Return original data only when all boxes are removed by warping.
if isempty(indices)
    B = A;
end
end

[Scot Bishop]

The modified functions were extremely helpful! Thank you Michael!

[ahmed shahin]

thanks you it is working now

[Jeremy Bulleid]

Thanks Michael, I had a similar problem and your solution works well.

[庆 王]

thanks a lot , it works

[Gyubin Hwang]

you saved the day. thank you!

[泽宇 王]

Thank you, sir.

[Vladimir Socha]

cool thx

[userlifeline]

THX!!!

Answer 3

[Samuel Manickavasagam S]

Tejas Phutane i too have this error

[TEJAS PHUTANE]

I tried verifying 1000 images in batches of 50 images and found errors in total 60 images.This method solved my problem