IFF SDK benchmarks
Testing results for common Machine Vision application where IFF SDK helps:
Order IFF SDK test drive!
Single and Multi-camera acquisition, color processing and streaming with NVIDIA Jetson Orin
Here are the results of testing the MRTech IFF SDK on machine vision systems equipped with XIMEA PCIe cameras and NVIDIA Jetson modules. Among other things, the latency of glass-to-glass image processing was measured when images from one or more XIMEA PCIe cameras (connected to a Jetson Orin module) were transmitted to a receiving computer.
Glass-to-glass latency refers to the time delay between light hitting the camera sensor and the resulting image being displayed on the viewer’s screen.
Hardware setup
- One or several XIMEA PCIe cameras from xiX and xiX-Xtreme camera series.
- NVIDIA Jetson modules:
- Orin NX with XIMEA XEC-NX-3P-X2G3 carrier board
- AGX Orin with Forge carrier board from Connect Tech
- MSI laptop with NVIDIA GPU and 240Hz display as the receiving computer
- LatencyKit from Science Mosaic to measure glass-to-glass latency
Software configuration for the transmitting side
IFF SDK ‘farsight’ sample application with the following processing pipeline:
- acquisition from XIMEA camera
- color pre-processing on GPU:
- black level subtraction
- histogram calculation
- white balance
- demosaicing
- color correction
- gamma
- image format conversion
- automatic control of exposure time and white balance
- H.265 encoding
- RTSP streaming over the network
Click to enlarge!
Software on the receiving side
IFF SDK ‘imagebroker’ sample application to receive, decode image stream and then export and render the images on the laptop screen using OpenCV library.
The following configuration and processing parameters are presented below:
- Camera configuration
-
Camera interface
- Image size (width x height in pixels)
-
Camera framerate for 12-bit raw image mode
- Image processing time on the remote computer / Glass-to-Glass latency from the camera to the receiving computer screen
-
Power consumption of NVIDIA Jetson module
The results of the Glass-to-glass latency measurements for each case are shown in the figures.
NVIDIA Jetson Orin NX
One camera
PCIe X4G3
Full HD (1920×1080)
390 FPS
9 ms / 28 ms
41 W
Click to enlarge!
One camera
PCIe X4G3
4K (3840×2160)
109 FPS
24 ms / 50 ms
42 W
Two cameras
2x PCIe X2G2
2x 8.1 MP (2848×2848)
2x 55 FPS
35 ms / 69 ms
37 W
Three cameras
PCIe X2G2
3x Full HD (1920×1080)
3x 120 FPS
13 ms / 43 ms
43 W
NVIDIA Jetson AGX Orin
One camera
PCIe X4G3
Full HD (1920×1080)
390 FPS
6 ms / 24 ms
44 W
Click to enlarge!
One camera
PCIe X4G3
4K (3840×2160)
140 FPS
17 ms / 42 ms
51 W
Two cameras
2x PCIe X2G2
2x 8.1 MP (2848×2848)
2x 60 FPS
25 ms / 68 ms
50 W
Explore all the IFF SDK benchmarks for NVIDIA Jetson Orin NX and AGX Orin
Click here and contact us to unlock your system potential!
Streaming of 8K UHD (7680×4320) images
The testing results when streaming 8K images from various host computers are presented.
Configuration and image processing parameters:
- Host computer
- Camera framerate for 12-bit raw image mode
- Image processing time on the remote computer / Glass-to-Glass latency from the camera to the receiving computer screen
- Power consumption for NVIDIA Jetson module
Click to enlarge!
Jetson AGX Orin
35 FPS
79 ms / 115 ms
47 W
Jetson Orin NX
17 FPS
136 ms / 175 ms
26 W
x86 host with NVIDIA RTX 2080Ti
50 FPS
30 ms / 90 ms
Denoise
To remove noise from images, IFF SDK has a denoising module that uses Discrete Wavelet Transform (DWT) and thresholding. There are two denoise options available in the SDK:
- RGB denoise which is applied after demosaicing (debayering),
- RAW denoise for Mono and Bayer (raw) images.
See denoise implementation details and parameter descriptions in the IFF SDK technical manual.
Testing results of the denoising module on a system with 4K camera and different NVIDIA GPUs are shown in the table.
Denoise performance for 4K UHD (3840 x 2160) 16-bit images
GeForce RTX 3070 Ti Laptop
- RAW Denoise: 1.77 ms
- RGB Denoise: 3.25 ms
GeForce RTX 2080 Ti
- RAW Denoise: 2.20 ms
- RGB Denoise: 3.55 ms
Jetson AGX Orin
- RAW Denoise: 3.72 ms
- RGB Denoise: 7.19 ms
Jetson AGX Xavier
- RAW Denoise: 6.76 ms
- RGB Denoise: 13.45 ms
Image comparison
Download the original TIFF images to explore the noise reduction result for a specific case with XIMEA MC124CG-SY color camera as an image source and laptop equipped with RTX 3070 Ti GPU.
Standard image processing pipeline with:
- 3840 x 2160 ROI on the camera
- DFPD high-quality demosaic with 11×11 processing window
- Denoise:
- without any denoise or
- with both RAW and RGB denoise options applied
- Image recording in TIFF format
Without denoise
After RAW and RGB denoise
Comparison of two cropped (900 x 600) images with Before/After image slider:
- Before – without any denoise
- After – with both RAW and RGB denoise
On-demand IFF SDK testing for your application
Submit your request for testing
