Samples - Object Detection
Samples - Object Detection
This section contains samples demonstrating how to use the ZED SDK’s Object Detection module to detect, classify and locate objects in 3D space.
This module requires a stereo camera equipped with an inertial sensor (IMU), such as the ZED 2/2i, ZED Mini, ZED X, ZED X Mini, ZED X Nano. The original ZED is not supported.
Getting Started
- First, download the latest version of the ZED SDK.
- Download the Object Detection sample code in C++, Python or C#.
Samples in this category
Image Viewer
Displays detected objects in a 2D view, making it simple to identify and track objects of interest over the live image. (C++, Python, C#)
Bird’s Eye Viewer
Presents detected objects in a top-down view alongside the 3D point cloud, giving an intuitive perspective on where objects are placed in the scene. (C++, Python, C#)
Concurrent Detections
Shows how to run the Object Detection and Body Tracking modules simultaneously in a single application. (C++, Python)
Custom Detector
Shows how to feed your own custom 2D object detector (YOLO/Ultralytics models, in ONNX, TensorRT or OpenCV DNN format) into the ZED SDK, which then computes 3D localization, bounding boxes and tracking on top of your detections. Multiple implementation variants are provided, covering different model formats and inference backends. (C++, Python)
Multi-Camera
Shows how to use the Fusion API to combine object detections from multiple cameras covering the same space, for more robust 3D detection and tracking. (C++)
Multi-Camera, Multi-Model
Shows how to run different detection models on different cameras and fuse the resulting detections through the Fusion API. (C++)