nrparikh's profile - activity

The project I am working on has a requirement to detect a person in the frame at a weird angle. I have tried using cascade classifiers but they don't work very well in my case. As a solution to this, I want to use TensorFlow's object detection API specifically import inception model into OpenCV and then do all the processing in OpenCV itself. I have used GOTURN tracker before which uses dnn module of OpenCV and reads the caffee model. I want to do something very similar, has anyone done anything similar to this or can anyone point me in the right direction?

Thank you for the quick reply...I did some digging into it and as far as I understand I can import models into OpenCV wh

Thanks...the pointer was really helpful! I have one question though do you know which model has human head/upper body d

Can we use pretrained TensorFlow model to detect objects in OpenCV? The project I am working on has a requirement to det

I tried your files and it works well with your files. I found the reason which might probably be the reason causing the error. The unzipping of the ".caffemodel" files was not proper and consequently, the creation of binary file might not be proper. When I did it again and concatenated the files it worked with those files also.

Thank you so much for your help!

The method suggested by 00zetti is correct. If still facing issues, please ensure following points:

• You have concatenated the zip files into single file named "goturn.caffemodel"
• You need to paste both files i.e. "goturn.protoxt" and "goturn.caffemodel" in your working directory. You can check your working directory using getcwd() command.

Credits to LBerger.

I tried building only tracking module but it seems like it didn't work properly. So, I built the library again and it works after rebuilding the whole library. And as you mentioned I got an error:

`OpenCV Error: Assertion Failed ((bias && l->blobs.size() == 2) || (!bias && l->blobs.size() == 1)) in initConvDeconvLayerFromCaffe, file /../opencv/modules/dnn/src/layers/convolution_layer.cpp, line 1204 terminate called after throwing an instance of 'cv::Exception'

what(): /../opencv/modules/dnn/src/layers/convolution_layer.cpp:1204: error:(-215) (bias && l->blobs.size() == 2) || (!bias && l->blobs.size() == 1) in function initConvDeconvLayerFromCaffe

Aborted (core dumped)`

I again went through the link you posted but it seems like there is no solution to this error, is it?

I'll need to build and install the library again after changing the lines, am I correct?

I changed the folder in which the files were and OpenCV is able to open the files but now I am getting another error: "Requested object was not found (Requested blob ".data1" not found in setInput, file /.../src/dnn.cpp".

I went through the link you posted earlier but couldn't find the solution, do you know how to solve it?

Yes, I have "Win" if I give the full path to the file but "Lost" if I only mention the name of the file. Please note that both the files are in the same folder. The gist is if I use test script to open the file and give the full path to the file, it will open but if I add the tracker in it and then OpenCV tries to open the file it fails.

My question is that do I need to specify the path of the file? If yes then how do I let OpenCV know the path?

I did just as you mentioned and it results in failure. I have ensured that permissions are correct but still, somehow the executable fails to open the file. Do you have any insight to it? I tried giving the full path to the file, it works but the question is how do I specify the path in OpenCV API?

I added to the directory as you mentioned but it still fails with same error. Additionally, I also tried to make it run by copying it to the location of caffe_io.cpp but failed again. Is there anything that I am missing?

Yes, I am pretty sure that I have both the files needed i.e. goturn.prototxt and goturn.caffemodel.

By executable file (the file being caffe_io.cpp) you mean the source file or the executable file generated after compilation?

I am using OpenCV 3.3.0 with C++ on Ubuntu 14.04. While trying to use GOTURN tracker, I am getting an error "Can't open 'goturn.prototxt' file in ReadProtoFromTextFile". Please note that I have done following steps:

• Clone opencv, opencv_contrib repository
• Build the library as mentioned on the official page
• Follow the steps mentioned here

Additionally, I have also tried to build the library after cloning opencv_extra repository and setting OPENCV_TEST_DATA_PATH to opencv_extra/testdata but still, it fails (I couldn't find the caffemodel in latest branch). I cloned this repository and linked it using OPENCV_TEST_DATA_PATH and I am still getting the same error.

Also, note that I have ensured the user has all required permissions.

Please let me know if I am missing any steps while building or installing the library? If someone who has experienced this kind of error can provide steps or link the tutorial, I will really appreciate it!