Preserving Image Quality at 10:1+ Compression
- Maintains image quality and original SNR while improving compression ratios
- Replaces unnecessary noise with compression-efficient data
- Improved compression for visually lossless applications
- On-the-fly adaptive training optimized for each video stream
- Compression exceeding 1.2 GPixels/s with complete CPU offload on the frame grabber’s FPGA
- Less than one frame period of latency for real-time responsiveness
- Integrated into Gidel’s GIL FPGA image processing flow
- Optimized for static cameras
FPGA Image Compression for High-Bandwidth Imaging
Gidel’s Quality+ is a proprietary FPGA image compression IP developed as a high-performance alternative to traditional lossless compression.
Quality+ performs real-time FPGA-based compression for Color Filter Array formats, such as Bayer, as well as monochrome and RGB images and videos. Designed for high-speed, high-bandwidth imaging, Quality+ achieves compression ratios up to 10:1+ while preserving the original image quality.
Unlike conventional visually lossless methods that compromise fidelity, Quality+ maintains the original signal-to-noise ratio, SNR, by replacing unnecessary noise with compression-efficient data. This maximizes bandwidth efficiency without degrading image quality.
Adaptive FPGA Compression with On-the-Fly Training
With on-the-fly adaptive training, Quality+ continuously optimizes compression for each video stream in real time. Operating entirely on the FPGA, it removes CPU and GPU load while supporting throughput exceeding 1.2 GPixels/s with less than one frame period of latency.
Ideal for high-bandwidth Imaging & Vision Systems, including medical imaging, surveillance, and defense applications, Quality+ reduces data rates and storage requirements while preserving image integrity in real time.
Related Products
General
- Color Filter Array (e.g., Bayer), RGB, Monochrome images.
- For other image formats, contact Gidel.
- FantoVision Edge Computers
- PCIe Frame Grabbers
- PCIe FPGA accelerators (For RAW data)
- Ultra-Compact FPGA modules
- Single frame
- Video stream
Gidel’s FPGA Quality+ compression is highly modular.
Looking to adapt it to your application requirements?
| Name | Description | Type | Size | |
|---|---|---|---|---|
|
Quality+ Compression
|
Datasheet | 163.7 KB | ||
|
Quality+ Compression
|
System Block Diagram: Compression in the FPGA data flow | PNG | 55.3 KB | |
|
Quality+ Compression
|
Performance Table | PNG | 5.2 KB | |
|
Video: Image processing acceleration on NVIDIA Jetson embedded computer
|
Demo at the Embedded World 2024 show | Link | — | Watch |
|
Compressing image data, not image quality
|
As today’s cameras deliver higher resolutions and frame rates and new systems architectures such as... | Link | — | Read Article |
Related Videos
FAQ
Traditional lossless compression often struggles to achieve high ratios with noisy, high-resolution data. Quality+ uses a proprietary FPGA-based algorithm that maintains the original Signal-to-Noise Ratio, SNR, by replacing unnecessary noise with compression-efficient data. This allows for significantly higher compression ratios, up to 10:1 or more, while preserving the image integrity required for advanced analysis and vision.
The FPGA IP features on-the-fly adaptive training, which means it continuously optimizes the compression parameters for each specific video stream in real time. This ensures that the compression is always perfectly tuned to the current visual data, providing an improved version for visually lossless applications without the artifacts found in standard methods.
The Quality+ compression IP can be deployed across Gidel FPGA platforms, including FantoVision Mini Edge AI systems, High-Bandwidth PCIe Frame Grabbers, FPGA Compute Accelerators, and Ultra-Compact FPGA Modules.
Quality+ is designed for massive data volumes, supporting throughput exceeding 1.2 GPixels per second. Because it operates entirely on the FPGA, it offers near-zero latency, less than one frame period, and provides total CPU offload, allowing the host system to focus on high-level analytics rather than data handling.
While Quality+ delivers high-performance FPGA compression across many scenarios, it is specifically optimized for static cameras. In these environments, it excels at identifying and maintaining high-fidelity detail over long-duration recordings or real-time streams, making it ideal for surveillance, medical imaging, and defense.
Quality+ supports Color Filter Array formats, such as Bayer, as well as monochrome and RGB images. The FPGA compression IP supports both single-frame capture and continuous video streams, ensuring compatibility with triggered inspection and constant monitoring workflows.