RGB & MobileNetSSD @ 4K

This example shows how to run MobileNetv2SSD on the RGB input frame, and how to display both the RGB preview and the metadata results from the MobileNetv2SSD on the preview. The preview size is set to 4K resolution.

It’s a variation of RGB & MobilenetSSD.

Similiar samples:

Demo

Setup

Please run the install script to download all required dependencies. Please note that this script must be ran from git context, so you have to download the depthai-python repository first and then run the script

git clone https://github.com/luxonis/depthai-python.git
cd depthai-python/examples
python3 install_requirements.py

For additional information, please follow installation guide

This example also requires MobilenetSDD blob (mobilenet-ssd_openvino_2021.2_5shave.blob file) to work - you can download it from here

Source code

Also available on GitHub

  1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
#!/usr/bin/env python3

from pathlib import Path
import sys
import cv2
import depthai as dai
import numpy as np

# Get argument first
nnPath = str((Path(__file__).parent / Path('models/mobilenet-ssd_openvino_2021.4_5shave.blob')).resolve().absolute())
if len(sys.argv) > 1:
    nnPath = sys.argv[1]

if not Path(nnPath).exists():
    import sys
    raise FileNotFoundError(f'Required file/s not found, please run "{sys.executable} install_requirements.py"')

# MobilenetSSD label texts
labelMap = ["background", "aeroplane", "bicycle", "bird", "boat", "bottle", "bus", "car", "cat", "chair", "cow",
            "diningtable", "dog", "horse", "motorbike", "person", "pottedplant", "sheep", "sofa", "train", "tvmonitor"]

# Create pipeline
pipeline = dai.Pipeline()

# Define sources and outputs
camRgb = pipeline.createColorCamera()
nn = pipeline.createMobileNetDetectionNetwork()

xoutVideo = pipeline.createXLinkOut()
xoutPreview = pipeline.createXLinkOut()
nnOut = pipeline.createXLinkOut()

xoutVideo.setStreamName("video")
xoutPreview.setStreamName("preview")
nnOut.setStreamName("nn")

# Properties
camRgb.setPreviewSize(300, 300)    # NN input
camRgb.setResolution(dai.ColorCameraProperties.SensorResolution.THE_4_K)
camRgb.setInterleaved(False)
camRgb.setPreviewKeepAspectRatio(False)
# Define a neural network that will make predictions based on the source frames
nn.setConfidenceThreshold(0.5)
nn.setBlobPath(nnPath)
nn.setNumInferenceThreads(2)
nn.input.setBlocking(False)

# Linking
camRgb.video.link(xoutVideo.input)
camRgb.preview.link(xoutPreview.input)
camRgb.preview.link(nn.input)
nn.out.link(nnOut.input)

# Connect to device and start pipeline
with dai.Device(pipeline) as device:

    # Output queues will be used to get the frames and nn data from the outputs defined above
    qVideo = device.getOutputQueue(name="video", maxSize=4, blocking=False)
    qPreview = device.getOutputQueue(name="preview", maxSize=4, blocking=False)
    qDet = device.getOutputQueue(name="nn", maxSize=4, blocking=False)

    previewFrame = None
    videoFrame = None
    detections = []

    # nn data, being the bounding box locations, are in <0..1> range - they need to be normalized with frame width/height
    def frameNorm(frame, bbox):
        normVals = np.full(len(bbox), frame.shape[0])
        normVals[::2] = frame.shape[1]
        return (np.clip(np.array(bbox), 0, 1) * normVals).astype(int)

    def displayFrame(name, frame):
        color = (255, 0, 0)
        for detection in detections:
            bbox = frameNorm(frame, (detection.xmin, detection.ymin, detection.xmax, detection.ymax))
            cv2.putText(frame, labelMap[detection.label], (bbox[0] + 10, bbox[1] + 20), cv2.FONT_HERSHEY_TRIPLEX, 0.5, color)
            cv2.putText(frame, f"{int(detection.confidence * 100)}%", (bbox[0] + 10, bbox[1] + 40), cv2.FONT_HERSHEY_TRIPLEX, 0.5, color)
            cv2.rectangle(frame, (bbox[0], bbox[1]), (bbox[2], bbox[3]), color, 2)
        # Show the frame
        cv2.imshow(name, frame)

    cv2.namedWindow("video", cv2.WINDOW_NORMAL)
    cv2.resizeWindow("video", 1280, 720)
    print("Resize video window with mouse drag!")

    while True:
        # Instead of get (blocking), we use tryGet (nonblocking) which will return the available data or None otherwise
        inVideo = qVideo.tryGet()
        inPreview = qPreview.tryGet()
        inDet = qDet.tryGet()

        if inVideo is not None:
            videoFrame = inVideo.getCvFrame()

        if inPreview is not None:
            previewFrame = inPreview.getCvFrame()

        if inDet is not None:
            detections = inDet.detections

        if videoFrame is not None:
            displayFrame("video", videoFrame)

        if previewFrame is not None:
            displayFrame("preview", previewFrame)

        if cv2.waitKey(1) == ord('q'):
            break

Also available on GitHub

  1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
#include <iostream>

// Inludes common necessary includes for development using depthai library
#include "depthai/depthai.hpp"

// MobilenetSSD label texts
static const std::vector<std::string> labelMap = {"background", "aeroplane", "bicycle",     "bird",  "boat",        "bottle", "bus",
                                                  "car",        "cat",       "chair",       "cow",   "diningtable", "dog",    "horse",
                                                  "motorbike",  "person",    "pottedplant", "sheep", "sofa",        "train",  "tvmonitor"};

int main(int argc, char** argv) {
    using namespace std;
    // Default blob path provided by Hunter private data download
    // Applicable for easier example usage only
    std::string nnPath(BLOB_PATH);

    // If path to blob specified, use that
    if(argc > 1) {
        nnPath = std::string(argv[1]);
    }

    // Print which blob we are using
    printf("Using blob at path: %s\n", nnPath.c_str());

    // Create pipeline
    dai::Pipeline pipeline;

    // Define sources and outputs
    auto camRgb = pipeline.create<dai::node::ColorCamera>();
    auto nn = pipeline.create<dai::node::MobileNetDetectionNetwork>();

    auto xoutVideo = pipeline.create<dai::node::XLinkOut>();
    auto xoutPreview = pipeline.create<dai::node::XLinkOut>();
    auto nnOut = pipeline.create<dai::node::XLinkOut>();

    xoutVideo->setStreamName("video");
    xoutPreview->setStreamName("preview");
    nnOut->setStreamName("nn");

    // Properties
    camRgb->setPreviewSize(300, 300);  // NN input
    camRgb->setResolution(dai::ColorCameraProperties::SensorResolution::THE_4_K);
    camRgb->setInterleaved(false);
    camRgb->setPreviewKeepAspectRatio(false);
    // Define a neural network that will make predictions based on the source frames
    nn->setConfidenceThreshold(0.5);
    nn->setBlobPath(nnPath);
    nn->setNumInferenceThreads(2);
    nn->input.setBlocking(false);

    // Linking
    camRgb->video.link(xoutVideo->input);
    camRgb->preview.link(xoutPreview->input);
    camRgb->preview.link(nn->input);
    nn->out.link(nnOut->input);

    // Connect to device and start pipeline
    dai::Device device(pipeline);

    // Output queues will be used to get the frames and nn data from the outputs defined above
    auto qVideo = device.getOutputQueue("video", 4, false);
    auto qPreview = device.getOutputQueue("preview", 4, false);
    auto qDet = device.getOutputQueue("nn", 4, false);

    cv::Mat previewFrame;
    cv::Mat videoFrame;
    std::vector<dai::ImgDetection> detections;

    // Add bounding boxes and text to the frame and show it to the user
    auto displayFrame = [](std::string name, cv::Mat frame, std::vector<dai::ImgDetection>& detections) {
        auto color = cv::Scalar(255, 0, 0);
        // nn data, being the bounding box locations, are in <0..1> range - they need to be normalized with frame width/height
        for(auto& detection : detections) {
            int x1 = detection.xmin * frame.cols;
            int y1 = detection.ymin * frame.rows;
            int x2 = detection.xmax * frame.cols;
            int y2 = detection.ymax * frame.rows;

            int labelIndex = detection.label;
            std::string labelStr = to_string(labelIndex);
            if(labelIndex < labelMap.size()) {
                labelStr = labelMap[labelIndex];
            }
            cv::putText(frame, labelStr, cv::Point(x1 + 10, y1 + 20), cv::FONT_HERSHEY_TRIPLEX, 0.5, color);
            std::stringstream confStr;
            confStr << std::fixed << std::setprecision(2) << detection.confidence * 100;
            cv::putText(frame, confStr.str(), cv::Point(x1 + 10, y1 + 40), cv::FONT_HERSHEY_TRIPLEX, 0.5, color);
            cv::rectangle(frame, cv::Rect(cv::Point(x1, y1), cv::Point(x2, y2)), color, cv::FONT_HERSHEY_SIMPLEX);
        }
        // Show the frame
        cv::imshow(name, frame);
    };

    cv::namedWindow("video", cv::WINDOW_NORMAL);
    cv::resizeWindow("video", 1280, 720);
    cout << "Resize video window with mouse drag!" << endl;

    while(true) {
        // Instead of get (blocking), we use tryGet (nonblocking) which will return the available data or None otherwise
        auto inVideo = qVideo->tryGet<dai::ImgFrame>();
        auto inPreview = qPreview->tryGet<dai::ImgFrame>();
        auto inDet = qDet->tryGet<dai::ImgDetections>();

        if(inVideo) {
            videoFrame = inVideo->getCvFrame();
        }

        if(inPreview) {
            previewFrame = inPreview->getCvFrame();
        }

        if(inDet) {
            detections = inDet->detections;
        }

        if(!videoFrame.empty()) {
            displayFrame("video", videoFrame, detections);
        }

        if(!previewFrame.empty()) {
            displayFrame("preview", previewFrame, detections);
        }

        int key = cv::waitKey(1);
        if(key == 'q' || key == 'Q') {
            return 0;
        }
    }
    return 0;
}

Got questions?

We’re always happy to help with code or other questions you might have.