cpp prediction example for video caffe

[Mohammad Moradi]

Hi, I'm trying to write a code to classify videos. I trained my model based on https://github.com/chuckcho/video-caffe and the results are nice. but I don't know how to read sequence of image and set them as input blob for running forward pass. I tried to change caffe cpp example for prediction(https://github.com/BVLC/caffe/blob/master/examples/cpp_classification/classification.cpp) but I don't understand how I should convert sequence of images to blob. Here is my deploy file that I want to use for prediction. Any hints would be appreciated.
Thanks in advance