01 / Environment review
Share the exact board, operating system, BSP, RKNN Runtime, memory and input format. Compatibility is evaluated against this environment, not a chipset name alone.
RK3588 SOLUTIONS
Integrate feature extraction into RK3588 devices.
We review your RK3588 Linux ARM64 environment, agree on evaluation targets, and define the SDK and integration work needed for deployment.
02 / Local processing
Your application decodes video or supplies image frames, calls the SDK and receives feature vectors. Image inference runs on the device. The application controls storage and network communication; license activation and any offline requirements are agreed separately.
03 / Measured validation
Before deployment, test the agreed number of video streams, frames processed per second per stream, input resolution, memory use and processing latency. Results apply to the tested hardware, model and settings; they do not imply an unlimited channel count.
04 / Production integration
Confirm the SDK version, license activation method, update policy and licensed product scope. Porting to another board or a substantially different BSP requires a separate assessment of work and fees.
Integration details
Specifications and scope
- Best fit
- NVR, camera and embedded AI device manufacturers; compatibility is reviewed for each target device
- Initial platform
- RK3588 on Linux ARM64
- Evaluation inputs
- Board details, software environment and workload targets
- Deliverables
- Agreed SDK package, integration samples and validation notes
Let’s build your next edge product.
Start with your hardware, your requirements and a focused evaluation.