Hello,
I am assuming that you are making use of a Neural Network for classification purpose.
In order to train Network using video, extract all the frames from each video and create an imageset for all frames. You may make use of HOG or SURF features from the frames and fitcecoc can be used to classify the frames using the description of each imageset as a label.
Following code may be implemented to make a row-wise matrix for frames using SURF features as descriptors.
for i=1:length(frames)
st=strcat('D:\test\chair\000',num2str(i,'%02d'),'.png'); %%Directory to image set
I = imread(st);
I = rgb2gray(I);
points = detectSURFFeatures(I);
[features, valid_points] = extractFeatures(I, points);
[s1,s2]=size(features);
S=s1*s2;
ftr=reshape(features,1,S);
F(i,:)=ftr;
end
Please refer to the following links:
- https://www.mathworks.com/help/deeplearning/gs/create-simple-deep-learning-classification-network.html (Basic image classification using a simple convolution Neural Network)
- https://www.mathworks.com/videos/what-is-deep-learning-toolbox--1535667599631.html (Video)
- https://in.mathworks.com/help/deeplearning/ug/classify-videos-using-deep-learning.html (Classify Videos using Deep Learning)
- https://in.mathworks.com/matlabcentral/fileexchange/74402-video-classification-using-lstm-lstm (Video Classification using LSTM)
- https://www.mathworks.com/help/vision/examples/image-category-classification-using-bag-of-features.html (Image category classification using a "bag of features" approach which uses SURF features and SVM classifier)
- https://in.mathworks.com/help/vision/ref/extracthogfeatures.html (HOG Feature)
- https://in.mathworks.com/help/stats/fitcecoc.html (fitcecoc)
Hope it helps!!
