Waveshare AR0144 2MP Stereo USB Camera Module w/ 52mm Baseline

The camera is UVC-compliant and works driver-free with standard imaging libraries. You can use it directly with OpenCV (cv2.VideoCapture), ROS (usb_cam package), MATLAB Image Acquisition Toolbox, or Python libraries like PyAV and SimpleCV. On Linux, it appears as a standard V4L2 device. For stereo processing, you will need to split the combined 2560x720 stream into two 1280x720 images using your chosen software framework.

Depth calculation requires stereo vision algorithms. First, calibrate your camera pair using a checkerboard pattern to obtain intrinsic and extrinsic parameters, including the precise 52mm baseline. Then, perform stereo rectification to align image rows. Finally, compute a disparity map using algorithms like Block Matching (BM) or Semi-Global Block Matching (SGBM) in OpenCV. Depth (Z) is calculated using the formula Z = (f * B) / d, where f is focal length, B is baseline (52mm), and d is disparity in pixels.

Global shutter sensors expose all pixels simultaneously, capturing an instantaneous snapshot of the scene. This eliminates the rolling shutter artifact common in cheaper CMOS sensors, where image rows are captured sequentially, causing skew with fast motion or vibrations. For synchronized stereo vision and depth mapping, a global shutter ensures both cameras capture the exact same moment in time, which is critical for accurate disparity calculation and 3D reconstruction in dynamic environments like moving robots or drones.

Yes, the camera is compatible with Raspberry Pi (Pi 4, Pi 5) and other SBCs with a USB 2.0 host port. On the Raspberry Pi, ensure your OS supports USB Video Class (UVC) devices. You may need to adjust the USB bandwidth allocation if using other high-bandwidth peripherals. For processing dual 720p streams in real-time, a Raspberry Pi 4 or higher with adequate RAM is recommended. Use the libcamera-vid tool or OpenCV's VideoCapture for access.