Convert NV12@1920x1080 to RGB@224x224

Model Overview

A preprocessing model to convert NV12 formatted images from 1920x1080 resolution to RGB format at 224x224 resolution.

The Convert NV12@1920x1080 to RGB@224x224 model is developed and optimized for the Synaptics Astra™ SL1680 processor NPU and SL1640 processor NPU.

Model Features

• Model Type: Image Processing
• Input Size: NV12@1920x1080
• Output Size: RGB@224x224

ℹ️ INFO: This model is ready to use on Synaptics Astra Machina boards. An NPU optimized version of the Convert NV12@1920x1080 to RGB@224x224 is installed in the Astra SDK Image.

Deployment on Synaptics Astra SL1600 Series

This particular model is compiled for Synaptics Astra SL1680 processor. You can find this model already pre-installed on Machina™ Dev kit with SL1680 processor.

You can also find the same model compiled for Synaptics Astra SL1640 processor pre-installed on Machina™ Dev kit with SL1640 processor.

Synaptics Astra Machina™ is Modular developer kit for Astra SL-Series of high-performance IoT processors with integrated Synaptics Veros™ wireless connectivity solution. Learn more here

Application binary

The synap_cli_ip command line application supports image processing models like Convert NV12@1920x1080 to RGB@224x224, including super-resolution.

Inputs:

• The converted synap model (.synap extension)
• Optionally, a region of interest in the image (if supported by the model)
• One or more raw images (nv12, nv21, rgb, bgr, bgra, gray, or bin formats)

Outputs:

• Processed image files named outimage<i>_<W>x<H>.<ext>, where:
  • <i> is the input file index
  • <W>x<H> are the output dimensions
  • <ext> is the output format (e.g., nv12 or rgb)

Command line usage on Astra SL1680 and SL1640:

MODELS=/usr/share/synap/models/

cd $MODELS/image_processing/preprocess/model/convert/clone_224x224/convert_nv12@1920x1080_rgb@224x224

synap_cli_ip -m model.synap ../../sample/ref_1920x1080.nv12

Example output on SL1680:

Input image: ../../sample/ref_1920x1080.nv12
Inference time: 1.27 ms
Writing output to file: outimage0_224x224.rgb

💡NOTE: Input images are resized to the network input tensor size, except for nv12, where input files must match the network’s expected dimensions.

💡NOTE: Use image_to_raw from the SyNAP toolkit to convert JPEG/PNG images to raw formats (e.g., nv12).

Performance on NPU

Processors	Inference Time (ms)
SL1680	1.27
SL1640	1.25

Optimize and Customize the model

Advanced users may wish to customize the source model and recompile it for the Synaptics Astra NPU. Please refer to the Bring Your Own Model section for more information.

For reference, the .synap format model provided on the firmware image was compiled for the Synaptics Astra NPU with the following .yaml settings:

security:
  secure: ${ENV:SYNAP_SECURITY_ENABLED}
  file: ../../../../../security.yaml
inputs:
  - preprocess:
      type: NV12
      size:
        - 1920
        - 1080

License

The source model is licensed under Apache License 2.0.

The compiled model for on-device deployment is covered under the Synaptics Astra EULA.

Learn More

• Synaptics AI Developer Zone: Get started with documentation, tutorials and resources for your Edge AI journey.
• Astra Support Portal: Connect with our engineering team and community.