bool ReadImageSequenceToVolumeDatum(const char* img_dir, const int start_frm, const int label,
const int length, const int height, const int width, const int sampling_rate, VolumeDatum* datum){
char fn_im[256];
cv::Mat img, img_origin;
char *buffer = NULL;
int offset = 0;
int channel_size = 0;
int image_size = 0;
int data_size = 0;
datum->set_channels(3);
datum->set_length(length);
datum->set_label(label);
datum->clear_data();
datum->clear_float_data();
offset = 0;
int end_frm = start_frm + length * sampling_rate;
for (int i=start_frm; i<end_frm; i+=sampling_rate){
sprintf(fn_im, "%s/%06d.jpg", img_dir, i);
if (height > 0 && width > 0) {
img_origin = cv::imread(fn_im, CV_LOAD_IMAGE_COLOR);
if (!img_origin.data) {
LOG(ERROR) << "Could not open or find file " << fn_im;
return false;
}
cv::resize(img_origin, img, cv::Size(width, height));
img_origin.release();
} else {
img = cv::imread(fn_im, CV_LOAD_IMAGE_COLOR);
}
if (!img.data){
LOG(ERROR) << "Could not open or find file " << fn_im;
return false;
}
if (i==start_frm){
datum->set_height(img.rows);
datum->set_width(img.cols);
image_size = img.rows * img.cols;
channel_size = image_size * length;
data_size = channel_size * 3;
buffer = new char[data_size];
}
for (int c=0; c<3; c++){
ImageChannelToBuffer(&img, buffer + c * channel_size + offset, c);
}
offset += image_size;
}
CHECK(offset == channel_size) << "wrong offset size" << std::endl;
datum->set_data(buffer, data_size);
delete []buffer;
return true;
}
bool ReadVideoToVolumeDatumHelperSafe(const char* filename, const int start_frm,
const int label, const int length, const int height, const int width,
const int sampling_rate, VolumeDatum* datum){
cv::VideoCapture cap;
cv::Mat img, img_origin;
char *buffer = NULL;
int offset = 0;
int channel_size = 0;
int image_size = 0;
int data_size = 0;
int use_start_frm = start_frm;
cap.open(filename);
if (!cap.isOpened()){
LOG(ERROR) << "Cannot open " << filename;
return false;
}
datum->set_channels(3);
datum->set_length(length);
datum->set_label(label);
datum->clear_data();
datum->clear_float_data();
int num_of_frames = cap.get(CV_CAP_PROP_FRAME_COUNT);
if (num_of_frames<length*sampling_rate){
LOG(INFO) << filename << " does not have enough frames; having "
<< num_of_frames;
return false;
}
if (start_frm < 0){
use_start_frm = caffe_rng_rand()%(num_of_frames-length*sampling_rate+1);
}
offset = 0;
CHECK_GE(use_start_frm, 0) << "start frame must be greater or equal to 0";
// Instead of random acess, do sequentically access (avoid key-frame issue)
// This will keep use_start_frm frames
int sequential_counter = 0;
while (sequential_counter < use_start_frm) {
cap.read(img_origin);
sequential_counter++;
}
int end_frm = use_start_frm + length * sampling_rate;
CHECK_LE(end_frm, num_of_frames) << "end frame must be less or equal to num of frames";
for (int i=use_start_frm; i<end_frm; i++){
if (sampling_rate > 1) {
// If sampling_rate > 1, purposely keep some frames
if ((i-use_start_frm) % sampling_rate !=0) {
cap.read(img_origin);
continue;
}
}
if (height > 0 && width > 0){
cap.read(img_origin);
if (!img_origin.data){
LOG(INFO) << filename << " has no data at frame " << i;
if (buffer!=NULL)
delete[] buffer;
return false;
}
cv::resize(img_origin, img, cv::Size(width, height));
}
else
cap.read(img);
if (!img.data){
LOG(ERROR) << "Could not open or find file " << filename;
if (buffer!=NULL)
delete[] buffer;
return false;
}
if (i==use_start_frm){
datum->set_height(img.rows);
datum->set_width(img.cols);
image_size = img.rows * img.cols;
channel_size = image_size * length;
data_size = channel_size * 3;
buffer = new char[data_size];
}
for (int c=0; c<3; c++){
ImageChannelToBuffer(&img, buffer + c * channel_size + offset, c);
}
offset += image_size;
}
CHECK(offset == channel_size) << "wrong offset size" << std::endl;
datum->set_data(buffer, data_size);
delete []buffer;
cap.release();
return true;
}
bool ReadVideoToVolumeDatum(const char* filename, const int start_frm, const int label,
const int length, const int height, const int width, const int sampling_rate, VolumeDatum* datum) {
cv::VideoCapture cap;
cv::Mat img, img_origin;
char *buffer;
int offset, channel_size, image_size, data_size;
int use_start_frm = start_frm;
cap.open(filename);
if (!cap.isOpened()) {
LOG(ERROR) << "Cannot open " << filename;
return false;
}
datum->set_channels(3);
datum->set_length(length);
datum->set_label(label);
datum->clear_data();
datum->clear_float_data();
int num_of_frames = cap.get(CV_CAP_PROP_FRAME_COUNT);
if (num_of_frames<length*sampling_rate) {
LOG(INFO) << "not enough frames; having " << num_of_frames;
return false;
}
if (start_frm < 0) {
use_start_frm = rand()%(num_of_frames-length*sampling_rate+1);
}
offset = 0;
CHECK_GE(use_start_frm, 0) << "start frame must be greater or equal to 0";
cap.set(CV_CAP_PROP_POS_FRAMES, use_start_frm);
int end_frm = use_start_frm + length * sampling_rate;
//CHECK_LE(end_frm, num_of_frames) << "end frame must less or equal to num of frames";
CHECK_LE(end_frm-1, num_of_frames) << "end frame-1 must less or equal to num of frames (filename=" << filename << ")";
for (int i=use_start_frm; i<end_frm; i+=sampling_rate) {
if (sampling_rate > 1)
cap.set(CV_CAP_PROP_POS_FRAMES, i);
if (height > 0 && width > 0) {
cap.read(img_origin);
if (!img_origin.data) {
LOG(INFO) << "No data at frame " << i << " in "<< filename;
return false;
}
cv::resize(img_origin, img, cv::Size(width, height));
}
else
cap.read(img);
if (!img.data) {
LOG(ERROR) << "Could not open or find file " << filename;
return false;
}
if (i==use_start_frm) {
datum->set_height(img.rows);
datum->set_width(img.cols);
image_size = img.rows * img.cols;
channel_size = image_size * length;
data_size = channel_size * 3;
buffer = new char[data_size];
}
for (int c=0; c<3; c++) {
ImageChannelToBuffer(&img, buffer + c * channel_size + offset, c);
}
offset += image_size;
}
CHECK(offset == channel_size) << "wrong offset size" << std::endl;
datum->set_data(buffer, data_size);
delete []buffer;
cap.release();
return true;
}