# Object Tracker Replay This example demonstrates running YOLOv6-nano object detection with the [ObjectTracker](https://docs.luxonis.com/software-v3/depthai/depthai-components/nodes/object_tracker.md) node on a replayed video using the ReplayVideo host node (or a live camera when requested). It displays per-object Track IDs and status; this demo does not compute XYZ coordinates. ## Demo This example requires the DepthAI v3 API, see [installation instructions](https://docs.luxonis.com/software-v3/depthai.md). ## Run Default (demo clip): ```python python object_tracker_replay.py ``` Use your own video file: ```python python object_tracker_replay.py -i /path/to/video.mp4 ``` Use the live camera instead of a file: ```python python object_tracker_replay.py --camera True ``` ## Pipeline ## Source code #### Python ```python #!/usr/bin/env python3 import cv2 import depthai as dai import time from pathlib import Path from argparse import ArgumentParser scriptDir = Path(__file__).resolve().parent examplesRoot = (scriptDir / Path('../')).resolve() # This resolves the parent directory correctly models = examplesRoot / 'models' videoPath = models / 'construction_vest.mp4' parser = ArgumentParser() parser.add_argument("-i", "--inputVideo", default=videoPath, help="Input video name") parser.add_argument("-c", "--camera", type=bool, help="Use camera as input", default=False) args = parser.parse_args() # Create pipeline with dai.Pipeline() as pipeline: # Define sources and outputs inputSource = None if args.camera: camRgb = pipeline.create(dai.node.Camera).build(dai.CameraBoardSocket.CAM_A) inputSource = camRgb else: replay = pipeline.create(dai.node.ReplayVideo) replay.setReplayVideoFile(Path(args.inputVideo)) inputSource = replay detectionNetwork = pipeline.create(dai.node.DetectionNetwork).build(inputSource, "yolov6-nano") objectTracker = pipeline.create(dai.node.ObjectTracker) detectionNetwork.setConfidenceThreshold(0.6) detectionNetwork.input.setBlocking(False) labelMap = detectionNetwork.getClasses() objectTracker.setDetectionLabelsToTrack([0]) # track only person # possible tracking types: ZERO_TERM_COLOR_HISTOGRAM, ZERO_TERM_IMAGELESS, SHORT_TERM_IMAGELESS, SHORT_TERM_KCF objectTracker.setTrackerType(dai.TrackerType.SHORT_TERM_IMAGELESS) # take the smallest ID when new object is tracked, possible options: SMALLEST_ID, UNIQUE_ID objectTracker.setTrackerIdAssignmentPolicy(dai.TrackerIdAssignmentPolicy.SMALLEST_ID) preview = objectTracker.passthroughTrackerFrame.createOutputQueue() tracklets = objectTracker.out.createOutputQueue() detectionNetwork.passthrough.link(objectTracker.inputTrackerFrame) detectionNetwork.passthrough.link(objectTracker.inputDetectionFrame) detectionNetwork.out.link(objectTracker.inputDetections) startTime = time.monotonic() counter = 0 fps = 0 color = (255, 255, 255) pipeline.start() while(pipeline.isRunning()): imgFrame = preview.get() track = tracklets.get() assert isinstance(imgFrame, dai.ImgFrame), "Expected ImgFrame" assert isinstance(track, dai.Tracklets), "Expected Tracklets" counter+=1 current_time = time.monotonic() if (current_time - startTime) > 1 : fps = counter / (current_time - startTime) counter = 0 startTime = current_time frame = imgFrame.getCvFrame() trackletsData = track.tracklets for t in trackletsData: roi = t.roi.denormalize(frame.shape[1], frame.shape[0]) x1 = int(roi.topLeft().x) y1 = int(roi.topLeft().y) x2 = int(roi.bottomRight().x) y2 = int(roi.bottomRight().y) try: label = labelMap[t.label] except: label = t.label cv2.putText(frame, str(label), (x1 + 10, y1 + 20), cv2.FONT_HERSHEY_TRIPLEX, 0.5, 255) cv2.putText(frame, f"ID: {[t.id]}", (x1 + 10, y1 + 35), cv2.FONT_HERSHEY_TRIPLEX, 0.5, 255) cv2.putText(frame, t.status.name, (x1 + 10, y1 + 50), cv2.FONT_HERSHEY_TRIPLEX, 0.5, 255) cv2.rectangle(frame, (x1, y1), (x2, y2), color, cv2.FONT_HERSHEY_SIMPLEX) cv2.putText(frame, "NN fps: {:.2f}".format(fps), (2, frame.shape[0] - 4), cv2.FONT_HERSHEY_TRIPLEX, 0.4, color) cv2.imshow("tracker", frame) if cv2.waitKey(1) == ord('q'): break ``` #### C++ ```cpp #include #include #include #include "depthai/pipeline/datatype/Tracklets.hpp" int main() { // Create pipeline dai::Pipeline pipeline; // Define sources and outputs auto replay = pipeline.create(); replay->setReplayVideoFile(VIDEO_PATH); // Create spatial detection network dai::NNModelDescription modelDescription{"yolov6-nano"}; auto detectionNetwork = pipeline.create()->build(replay, modelDescription); detectionNetwork->setConfidenceThreshold(0.6f); detectionNetwork->input.setBlocking(false); // Create object tracker auto objectTracker = pipeline.create(); objectTracker->setDetectionLabelsToTrack({0}); // track only person objectTracker->setTrackerType(dai::TrackerType::SHORT_TERM_IMAGELESS); objectTracker->setTrackerIdAssignmentPolicy(dai::TrackerIdAssignmentPolicy::SMALLEST_ID); // Create output queues auto preview = objectTracker->passthroughTrackerFrame.createOutputQueue(); auto tracklets = objectTracker->out.createOutputQueue(); // Link nodes detectionNetwork->passthrough.link(objectTracker->inputTrackerFrame); detectionNetwork->passthrough.link(objectTracker->inputDetectionFrame); detectionNetwork->out.link(objectTracker->inputDetections); // Start pipeline pipeline.start(); // FPS calculation variables auto startTime = std::chrono::steady_clock::now(); int counter = 0; float fps = 0; cv::Scalar color(255, 255, 255); while(pipeline.isRunning()) { auto imgFrame = preview->get(); auto track = tracklets->get(); counter++; auto currentTime = std::chrono::steady_clock::now(); auto elapsed = std::chrono::duration_cast(currentTime - startTime).count(); if(elapsed >= 1) { fps = counter / static_cast(elapsed); counter = 0; startTime = currentTime; } cv::Mat frame = imgFrame->getCvFrame(); auto trackletsData = track->tracklets; for(const auto& t : trackletsData) { auto roi = t.roi.denormalize(frame.cols, frame.rows); int x1 = static_cast(roi.topLeft().x); int y1 = static_cast(roi.topLeft().y); int x2 = static_cast(roi.bottomRight().x); int y2 = static_cast(roi.bottomRight().y); std::string label; try { label = detectionNetwork->getClasses().value()[t.label]; } catch(...) { label = std::to_string(t.label); } cv::putText(frame, label, cv::Point(x1 + 10, y1 + 20), cv::FONT_HERSHEY_TRIPLEX, 0.5, color); cv::putText(frame, "ID: " + std::to_string(t.id), cv::Point(x1 + 10, y1 + 35), cv::FONT_HERSHEY_TRIPLEX, 0.5, color); cv::putText(frame, std::string(t.status == dai::Tracklet::TrackingStatus::LOST ? "LOST" : "TRACKED"), cv::Point(x1 + 10, y1 + 50), cv::FONT_HERSHEY_TRIPLEX, 0.5, color); cv::rectangle(frame, cv::Point(x1, y1), cv::Point(x2, y2), color, cv::FONT_HERSHEY_SIMPLEX); } cv::putText(frame, "NN fps: " + std::to_string(fps).substr(0, 4), cv::Point(2, frame.rows - 4), cv::FONT_HERSHEY_TRIPLEX, 0.4, color); cv::imshow("tracker", frame); if(cv::waitKey(1) == 'q') { break; } } return 0; } ``` ### Need assistance? Head over to [Discussion Forum](https://discuss.luxonis.com/) for technical support or any other questions you might have.