https://wiki.jw-tech.eu/index.php?action=history&feed=atom&title=WSAL22642_%E2%80%93_OAK-D-S2%2FenWSAL22642 – OAK-D-S2/en - Versionsgeschichte2026-10-23T21:51:54ZVersionsgeschichte dieser Seite in JW Tech WikiMediaWiki 1.43.9https://wiki.jw-tech.eu/index.php?title=WSAL22642_%E2%80%93_OAK-D-S2/en&diff=2416&oldid=prevJWTechPublisher: Freigegebene technische Dokumentation veröffentlicht; Text-only und noindex2026-09-06T06:04:14Z<p>Freigegebene technische Dokumentation veröffentlicht; Text-only und noindex</p>
<p><b>Neue Seite</b></p><div>__NOINDEX__<br />
'''Documentation notice:''' This technical documentation was independently rewritten from an archived supplier source. Check technical details and commands against the product version and manufacturer documentation before use.<br />
<br />
== Summary ==<br />
The OAK-D-S2 is a compact AI vision kit featuring an Intel Myriad X chip, a 12MP color camera, and binocular depth cameras. It supports Python and C++ development as well as H.265 streaming.<br />
<br />
== Related products at JW Tech ==<br />
* [https://shop.jw-tech.eu/p/WSAL22642 WSAL22642 – OAK-D-S2, 12MP, OpenCV AI Machine Vision Kit, Depth Measuring, Image Recognition]<br />
<br />
== Technical overview ==<br />
* Based on the OAK-D but more compact and lighter; the 5V power port present in the OAK-D has been removed.<br />
* Color camera: IMX378 sensor, 12 MP (4056 × 3040), 60 fps, autofocus from 8 cm to infinity.<br />
* Binocular monochrome cameras: OV9282 sensors, 1 MP each (1280 × 800), 120 fps, fixed focus from 50 cm or 19.6 cm.<br />
* Depth range: 0.7–12 m; theoretically up to 35 m, with about 6–10% measurement error at 30 m.<br />
* AI processor: Intel Myriad X with 4 TOPS of compute performance.<br />
* Video output: 4K/30 fps H.265 stream.<br />
* Interface: USB 3.0 Type-C for connection to computers or embedded hosts.<br />
* Supported development languages: Python and C++.<br />
* Metal enclosure; average power consumption approximately 2.5 W.<br />
* Supports common neural networks from the Caffe, TensorFlow, and MXNet platforms.<br />
<br />
'''Commands and examples – verify before use:''' <br />
<syntaxhighlight lang="text"><br />
git clone https://gitee.com/oakchina/depthai.git<br />
git clone https://gitee.com/oakchina/depthai-python.git<br />
echo 'SUBSYSTEM=="usb", ATTRS{idVendor}=="03e7", MODE="0666"' | sudo tee /etc/udev/rules.d/80-movidius.rules<br />
sudo udevadm control --reload-rules && sudo udevadm trigger<br />
python3 -m pip install -r depthai/requirements.txt -i https://pypi.tuna.tsinghua.edu.cn/simple<br />
python3 depthai/depthai_demo.py<br />
</syntaxhighlight><br />
<br />
== Notes and version constraints ==<br />
* The Linux installation steps refer to the depthai package; no standalone installation path for the OAK-D-S2 was specified in the source text—please verify the correct version of the depthai package before implementation.<br />
* The source text contains no explicit safety warnings; laser and radiation aspects must be checked separately.<br />
* Compatibility with certain hosts (e.g., Jetson) requires prior configuration of the target system to avoid conflicts with OpenCV.<br />
* Version-sensitive steps: Installation via Gitee mirrors may vary by region or network environment; please confirm repository accessibility.<br />
* No prices, warranties, or certifications were mentioned in the source text; such details must not be added.<br />
<br />
== Source and revision record ==<br />
* [https://www.waveshare.com/wiki/OAK-D-S2 OAK - Waveshare Wiki]<br />
* Source revision: 89738<br />
* Captured: 2026-09-03T04:14:25+00:00<br />
* SHA-256: <code>dcb2c8fa3c8b0cfe6539e80cb0d0f7176d53707fa529af4e33cb5d69d0966c36</code></div>JWTechPublisher