OpenCV fail to play video after start the client

[gupork]

Hi imsoo,

Thanks for the great sharing on this project. I have tried to build the project according to the instruction and the client can be run. However, once I start the following command: ./darknet_client -addr 127.0.0.1 -vid test_short.mp4 -out_vid.

Error message appear: OpenCV: FFMPEG: tag 0x5634504d/'MP4V' is not supported with codec id 13 and format 'mp4 / MP4 (MPEG-4 Part 14)' OpenCV: FFMPEG: fallback to use tag 0x7634706d/'mp4v'

The Input / Output window frozen on the first frame. The client can run but no action detected in the json file and the generated test_short_output.mp4 crash.

The client keep running without stop, when we stop it the error message is: (INPUT:13633): GLib-CRITICAL **: 13:53:56.806: Source ID 158 was not found when attempting to remove it

The following image is captured from several terminals of the deployed machine:

Please kindly provide some comments on how I begin the troubleshoot.

[imsoo]

@gupork Hi, Thank you for your interest in this project.

I think that there is a problem on the server side. darkent_client looks like fine.

:o: : Good, :x: : Fail Client --:o:--> Ventilator --:x:--> Worker --:x:--> Sink --:x:--> Client

First, please abort all process (Ventilator, Sink, Worker and Client) and retry it.

1. Run Server Processes (Sink, Ventilator, Worker)

./sink
./ventilator
./worker <CFG_PATH> <WEIGHTS_PATH> <NAMES_PATH> [-pose] [-gpu GPU_ID] [-thresh THRESH] 

2. Run Client

./darknet_client -addr 127.0.0.1 -vid test_short.mp4 -out_vid

3. Check that Ventilator ---> Worker ---> Sink Pipeline is working.

It should be like :

Ventilator | Recv From Client | SEQ :  0 | LEN : 12345
Ventilator | Send To Worker | SEQ :  0 | LEN : 12345
...
Worker | Recv From Ventilator | SEQ :  0 | LEN : 12345
Darknet | Detect | SEQ : 0 | Time : 50ms
Worker | Send To Sink | SEQ :  0 | LEN : 13456
...
Sink | Recv From Worker | SEQ : 0 | LEN : 13456
Sink | Pub To Client | SEQ : 0 | LEN : 13456

and then please let me know the result.

[gupork]

Just have a quick try: for the worker process, should I run two separate process: ./worker fight.cfg fight.weights coco.names -gpu 0 -thresh 0.2 ./worker openpose.cfg openpose.weight -gpu 0 -pose -thresh 0.2 Like the above?

If so, there should be 4 terminals running before the client process: ./sink ./ventilator ./worker fight.cfg fight.weights coco.names -gpu 0 -thresh 0.2 ./worker openpose.cfg openpose.weight -gpu 0 -pose -thresh 0.2

Right?

[gupork]

[imsoo]

if you want to two worker, use same command.

./worker openpose.cfg openpose.weight -gpu 0 -pose -thresh 0.2
./worker openpose.cfg openpose.weight -gpu 0 -pose -thresh 0.2

It should be like : ./sink ./ventilator ./worker openpose.cfg openpose.weight -gpu 0 -pose -thresh 0.2 ./worker openpose.cfg openpose.weight -gpu 0 -pose -thresh 0.2

[gupork]

So one worker is enough to make the whole pipeline work? Because I see the below instruction from the Readme seem two workers are required.

(Thanks again for your guidance. If I manage to run it would like to raise a cooperation project if you are interested!)

[gupork]

This time I have run: ./sink ./ventilator ./worker openpose.cfg openpose.weight -gpu 0 -pose -thresh 0.2 ./darknet_client -addr 127.0.0.1 -vid test_short.mp4 -out_vid

When it's running, both input / output window frozen:

The client manage to run but it won't stop and loop: R : 339 | C : 339 | F : 0 | T : 340 : 1 R : 339 | C : 339 | F : 0 | T : 340 : 1 R : 339 | C : 339 | F : 0 | T : 340 : 1 R : 339 | C : 339 | F : 0 | T : 340 : 1

When I check on the output video, its size: and fail to open it.

[imsoo]

So one worker is enough to make the whole pipeline work?

Right :+1:

If I manage to run it would like to raise a cooperation project if you are interested!

Thanks. always welcome :+1:

The client manage to run but it won't stop and loop:

Sorry, this client is unstable.

0. please abort sink and re-run sink.

1. change -out_vid to -out_json and check that json is fine.

./darknet_client -addr 127.0.0.1 -vid test_short.mp4 -out_json

2. if json is fine, change -out_json to -out_vid and retry it. (You have to re-run sink)

./darknet_client -addr 127.0.0.1 -vid test_short.mp4 -out_vid

if all data are received, try using ~~CTRL-Z~~ CTRL-C key to stop client program. ~~(check client stdout R : 339 | C : 339 | F : 0 | T : 340 : 1 R and C must same.)~~

3. if it doesn't work, please try using another video file. (You have to re-run sink)

[gupork]

Thanks for the tips! Finally the output window move for few seconds but stop still (the input window can run the whole video now)

The json can capture everything now.

However, when I get to out_vid, it comes out with this error:

I'm using a GPU Tesla T4 16GB on a local machine FYI.

[imsoo]

I think that we have a OpenCV problem.

please check that if using below command (not include -out_vid, -out_json), the input and output windows can play the whole video? (Please re-run all processes)

./darknet_client -addr 127.0.0.1 -vid test_short.mp4

and please let me know which version of openCV do you use.

---

Add : try this command.

./darknet_client -addr 127.0.0.1 -vid test_short.mp4 -out_vid -dont_show

[gupork]

I think that we have a OpenCV problem.

please check that if using below command (not include -out_vid, -out_json), the input and output windows can play the whole video? (Please re-run all processes)

./darknet_client -addr 127.0.0.1 -vid test_short.mp4

and please let me know which version of openCV do you use.

Add : try this command.

./darknet_client -addr 127.0.0.1 -vid test_short.mp4 -out_vid -dont_show

Taking out both output, the input window can play the whole video and the output window cannot.

Using dont_show, the output video file seem with normal size: However, it cannot be played via VLC player with error:

OpenCV version should be 4.2.0 by command:

However, given that I have tried multiple installation before, it appears to have 3.4.9 as well by following command:

[imsoo]

Sorry, I can't find what is the problem. I should test on my machine. I think it will take time to figure out what is the problem. Let me know if there's any progress.

[gupork]

Sorry, I can't find what is the problem. I should test on my machine. I think it will take time to figure out what is the problem. Let me know if there's any progress.

Thanks and understand. Do you have a recommended OpenCV version (other library as well)? If so I will try to set up the environment according to the library version you have.

I have uploaded the sample video here FYI: https://drive.google.com/file/d/19BHrzgebKcAHFHgw-2AukYmeTZM7XjWf/view?usp=sharing

Please let me know if you manage to run it. I shall send you an email to discuss in details for my project.

[imsoo]

@gupork Hi,

I found this issue.

Some UI backends may support "more or less" multi-thread interaction. But in general, multi-threading is not supported well by UI.

I think that win32 UI Backends support multi-thread but GTK may not.

So I moved GUI code(input_show_thread, output_show_thread) into main thread. It worked on my machine (Machine 1 : Ubuntu 14.04, OpenCV 3.4.0, Machine 2 : Windows 7, OpenCV 4.1.1)

Please try this and let me know about the result. I hope it works out.

Changed code (client/Main.cpp)

#include <zmq.h>
#include <iostream>
#include <thread>
#include <queue>
#include <chrono>
#include <thread>
#include <string>
#include <sstream>
#include <csignal>
#ifdef __linux__
#include <tbb/concurrent_priority_queue.h>
#elif _WIN32
#include <concurrent_priority_queue.h>
#endif
#include <opencv2/opencv.hpp>
#include "share_queue.hpp"
#include "frame.hpp"
#include "util.hpp"
#include "args.hpp"

#ifdef __linux__
#define FD_SETSIZE 4096
using namespace tbb;
#elif _WIN32
using namespace concurrency;
#endif
using namespace cv;
using namespace std;

// thread
void fetch_thread(void);
void capture_thread(void);
void recv_thread(void);
void output_show_thread(void);
void input_show_thread(void);

volatile bool fetch_flag = false;
volatile bool exit_flag = false;
volatile int final_exit_flag = 0;

// ZMQ
void *context;
void *sock_push;
void *sock_sub;

// pair
class ComparePair
{
public:
  bool operator()(pair<long, Frame> n1, pair<long, Frame> n2) {
    return n1.first > n2.first;
  }
};
Frame_pool *frame_pool;
concurrent_priority_queue<pair<long, Frame>, ComparePair> recv_queue;

// Queue
SharedQueue<Mat> cap_queue;
SharedQueue<Mat> fetch_queue;

// opencv
VideoCapture cap;
VideoWriter writer;
static Mat mat_show_output;
static Mat mat_show_input;
static Mat mat_recv;
static Mat mat_cap;
static Mat mat_fetch;

const int cap_width = 640;
const int cap_height = 480;
double delay;
long volatile show_frame = 1;
double end_frame;

// option
bool cam_input_flag;
bool vid_input_flag;
bool dont_show_flag;
bool json_output_flag;
bool vid_output_flag;

// output
string out_json_path;
string out_vid_path;

ofstream out_json_file;

void sig_handler(int s)
{
  exit_flag = true;
}

int main(int argc, char *argv[])
{
  if (argc < 2) {
    std::cerr << "Usage: " << argv[0] << " <-addr ADDR> <-cam CAM_NUM | -vid VIDEO_PATH> [-dont_show] [-out_json] [-out_vid] \n" << std::endl;
    return 0;
  }

  // install signal
  std::signal(SIGINT, sig_handler);

  // option init
  int cam_num = find_int_arg(argc, argv, "-cam", -1);
  if (cam_num != -1)
    cam_input_flag = true;

  const char *vid_def_path = "./test.mp4";
  const char *vid_path = find_char_arg(argc, argv, "-vid", vid_def_path);
  if (vid_path != vid_def_path)
    vid_input_flag = true;

  dont_show_flag = find_arg(argc, argv, "-dont_show");
  json_output_flag = find_arg(argc, argv, "-out_json");
  vid_output_flag = find_arg(argc, argv, "-out_vid");

  // frame_pool init
  frame_pool = new Frame_pool(5000);

  // ZMQ
  const char *addr = find_char_arg(argc, argv, "-addr", "127.0.0.1");
  context = zmq_ctx_new();

  sock_push = zmq_socket(context, ZMQ_PUSH);
  zmq_connect(sock_push, ((std::string("tcp://") + addr) + ":5575").c_str());

  sock_sub = zmq_socket(context, ZMQ_SUB);
  zmq_connect(sock_sub, ((std::string("tcp://") + addr) + ":5570").c_str());

  zmq_setsockopt(sock_sub, ZMQ_SUBSCRIBE, "", 0);

  if (vid_input_flag) {
    // VideoCaputre video
    cap = VideoCapture(vid_path);
    out_json_path = string(vid_path, strrchr(vid_path, '.')) + "_output.json";
    out_vid_path = string(vid_path, strrchr(vid_path, '.')) + "_output.mp4";
  }
  else if (cam_input_flag) {
    // VideoCapture cam
    cap = VideoCapture(cam_num);
    cap.set(CAP_PROP_FPS, 20);
    cap.set(CAP_PROP_BUFFERSIZE, 3);
    fetch_flag = true;
    out_json_path = "./cam_output.json";
    out_vid_path = "./cam_output.mp4";
  }
  else {
    // error
    std::cerr << "Usage: " << argv[0] << " <-addr ADDR> <-cam CAM_NUM | -vid VIDEO_PATH> [-dont_show] [-out_json] [-out_vid] \n" << std::endl;
    return 0;
  }

  if (!cap.isOpened()) {
    cerr << "Erro VideoCapture...\n";
    return -1;
  }

  double fps = cap.get(CAP_PROP_FPS);
  end_frame = cap.get(CAP_PROP_FRAME_COUNT);
  delay = cam_input_flag ? 1 : (1000.0 / fps);

  // read frame
  cap.read(mat_fetch);

  if (mat_fetch.empty()) {
    cerr << "Empty Mat Captured...\n";
    return 0;
  }

  mat_show_output = mat_fetch.clone();
  mat_show_input = mat_fetch.clone();

  // output init
  if (json_output_flag) {
    out_json_file = ofstream(out_json_path);
    if (!out_json_file.is_open()) {
      cerr << "output file : " << out_json_path << " open error \n";
      return 0;
    }
    out_json_file << "{\n \"det\": [\n";
  }

  if (vid_output_flag) {
    writer.open(out_vid_path, VideoWriter::fourcc('M', 'P', '4', 'V'), fps, Size(cap_width, cap_height), true);
    if (!writer.isOpened()) {
      cerr << "Erro VideoWriter...\n";
      return -1;
    }
  }

  // thread init
  thread thread_fetch(fetch_thread);
  thread_fetch.detach();

  while (!fetch_flag);

  thread thread_recv(recv_thread);
  thread thread_capture(capture_thread);

  thread_recv.detach();
  thread_capture.detach();

  if (!dont_show_flag) {
    cvNamedWindow("INPUT");
    moveWindow("INPUT", 30, 130);
    cv::imshow("INPUT", mat_show_input);

    cvNamedWindow("OUTPUT");
    moveWindow("OUTPUT", 670, 130);
    cv::imshow("OUTPUT", mat_show_output);
  }

  while (final_exit_flag)
  {
    // for debug
    cout << "R : " << recv_queue.size() << " | C : " << cap_queue.size() << " | F : " << fetch_queue.size() << " | T : " << end_frame << " : " << show_frame << endl;

    input_show_thread();
    output_show_thread();

    // wait key for exit
    if (waitKey(delay) >= 0)
      exit_flag = true;
  }

  cap.release();

  if (json_output_flag) {
    out_json_file << "\n ]\n}";
    out_json_file.close();
  }

  if (vid_output_flag) {
    writer.release();
  }

  delete frame_pool;
  zmq_close(sock_sub);
  zmq_close(sock_push);
  zmq_ctx_destroy(context);

  return 0;
}

#define FETCH_THRESH 100
#define FETCH_WAIT_THRESH 30
#define FETCH_STATE 0
#define FETCH_WAIT 1
void fetch_thread(void) {
  volatile int fetch_state = FETCH_STATE;
  final_exit_flag += 1;
  while (!exit_flag) {

    switch (fetch_state) {
    case FETCH_STATE:
      if (cap.grab()) {

        cap.retrieve(mat_fetch);

        // push fetch queue
        fetch_queue.push_back(mat_fetch.clone());

        // if cam dont wait
        if (!cam_input_flag && (fetch_queue.size() > FETCH_THRESH)) {
          fetch_state = FETCH_WAIT;
        }
      }
      // if fetch end
      else {
        final_exit_flag -= 1;
        return;
      }
      break;
    case FETCH_WAIT:
      fetch_flag = true;
      if (fetch_queue.size() < FETCH_WAIT_THRESH) {
        fetch_state = FETCH_STATE;
      }
      break;
    }
  }
  final_exit_flag -= 1;
}

void capture_thread(void) {
  static vector<int> param = { IMWRITE_JPEG_QUALITY, 50 };
  static vector<uchar> encode_buf(JSON_BUF_LEN);

  volatile int frame_seq_num = 1;
  string frame_seq;

  // for json
  unsigned char json_buf[JSON_BUF_LEN];
  int send_json_len;

  Frame frame = frame_pool->alloc_frame();

  final_exit_flag += 1;
  while (!exit_flag) {
    if (fetch_queue.size() < 1)
      continue;

    // get input mat 
    mat_cap = fetch_queue.front().clone();
    fetch_queue.pop_front();

    if (mat_cap.empty()) {
      cerr << "Empty Mat Captured...\n";
      continue;
    }

    // resize
    resize(mat_cap, mat_cap, Size(cap_width, cap_height));

    // push to cap queue (for display input)
    cap_queue.push_back(mat_cap.clone());

    // mat to jpg
    imencode(".jpg", mat_cap, encode_buf, param);

    // jpg to json (seq + msg)
    frame_seq = to_string(frame_seq_num);
    frame.seq_len = frame_seq.size();
    memcpy(frame.seq_buf, frame_seq.c_str(), frame.seq_len);

    frame.msg_len = encode_buf.size();
    memcpy(frame.msg_buf, &encode_buf[0], frame.msg_len);

    send_json_len = frame_to_json(json_buf, frame);

    // send json to server
    zmq_send(sock_push, json_buf, send_json_len, 0);

    frame_seq_num++;
  }
  frame_pool->free_frame(frame);
  final_exit_flag -= 1;
}

void recv_thread(void) {
  int recv_json_len;
  int frame_seq_num = 1;
  Frame frame;
  unsigned char json_buf[JSON_BUF_LEN];

  final_exit_flag += 1;
  while (!exit_flag) {
    recv_json_len = zmq_recv(sock_sub, json_buf, JSON_BUF_LEN, ZMQ_NOBLOCK);

    if (recv_json_len > 0) {
      frame = frame_pool->alloc_frame();
      json_buf[recv_json_len] = '\0';
      json_to_frame(json_buf, frame);

      frame_seq_num = str_to_int((const char *)frame.seq_buf, frame.seq_len);

      // push to recv_queue (for display output)
      pair<long, Frame> p = make_pair(frame_seq_num, frame);
      recv_queue.push(p);
    }
  }
  final_exit_flag -= 1;
}

#define DONT_SHOW 0
#define SHOW_START 1
#define DONT_SHOW_THRESH 2  // for buffering
#define SHOW_START_THRESH 1 // for buffering

int volatile show_state = DONT_SHOW;
void input_show_thread(void) {
  switch (show_state) {
  case DONT_SHOW:
    break;
  case SHOW_START:
    if (cap_queue.size() >= DONT_SHOW_THRESH) {
      mat_show_input = cap_queue.front().clone();
      cap_queue.pop_front();
    }
    break;
  }

  if (!dont_show_flag) {

    // draw text (INPUT) Left Upper corner  
    putText(mat_show_input, "INPUT", Point(10, 25),
      FONT_HERSHEY_SIMPLEX, 0.5, Scalar(0, 0, 255), 2);

    cv::imshow("INPUT", mat_show_input);
  }
}

void output_show_thread(void) {
  Frame frame;

  switch (show_state) {
  case DONT_SHOW:
    if (recv_queue.size() >= SHOW_START_THRESH) {
      show_state = SHOW_START;
    }
    break;
  case SHOW_START:
    if (recv_queue.size() >= DONT_SHOW_THRESH || (end_frame - show_frame) == 1) {
      pair<long, Frame> p;
      // try pop success
      while (1) {
        if (recv_queue.try_pop(p)) {
          // if right sequence
          if (p.first == show_frame) {

            frame = ((Frame)p.second);
            vector<uchar> decode_buf((unsigned char*)(frame.msg_buf), (unsigned char*)(frame.msg_buf) + frame.msg_len);

            // jpg to mat
            mat_show_output = imdecode(decode_buf, IMREAD_COLOR);

            // resize
            resize(mat_show_output, mat_recv, Size(cap_width, cap_height));

            // wirte out_json
            if (json_output_flag) {
              if (show_frame != 1)
                out_json_file << ",\n";
              out_json_file.write((const char*)frame.det_buf, frame.det_len);
            }

            // write out_vid
            if (vid_output_flag)
              writer.write(mat_show_output);

            // free frame
            frame_pool->free_frame(frame);
            show_frame++;
          }
          // wrong sequence
          else {
            recv_queue.push(p);
          }
          break;
        }
      }
    }
    else {
      show_state = DONT_SHOW;
    }
    break;
  }

  if (show_frame == end_frame)
    exit_flag = true;

  if (!dont_show_flag) {
    // draw text (OUTPUT) Left Upper corner  
    putText(mat_show_output, "OUTPUT", Point(10, 25),
      FONT_HERSHEY_SIMPLEX, 0.5, Scalar(0, 255, 0), 2);

    cv::imshow("OUTPUT", mat_show_output);
  }
}

[gupork]

@gupork Hi,

I found this issue.

Some UI backends may support "more or less" multi-thread interaction. But in general, multi-threading is not supported well by UI.

I think that win32 UI Backends support multi-thread but GTK may not.

So I moved GUI code(input_show_thread, output_show_thread) into main thread. It worked on my machine (Machine 1 : Ubuntu 14.04, OpenCV 3.4.0, Machine 2 : Windows 7, OpenCV 4.1.1)

Please try this and let me know about the result. I hope it works out.

Changed code (client/Main.cpp)

Hi imsoo!

I have no luck to run the "fight" detection unfortunately but I'm quite confident in your solution. Do you want to develop it further and apply on some real cases? If so, I would like to discuss with you in details via email or any convenient way.

[imsoo]

@gupork,

Sorry for my late reply. I sent it by email.

[yutao007]

the same problem to me .

[yutao007]

does it need stronger gpu ? i have used single gpu 1050 and 1060 .the trouble is same to you ! Communication is normal ，The Input / Output window is frozen on the first frame. The client can run but no action detected in the json file and the generated test_short_output.mp4 crash.

[yutao007]

last when se -dont_show it can generate outvideo. and when use small person video openpose and fight detection are not good