RFEL signal processing iq : fpga expertise

RFEL

Video Processing

IP Cores

Digital Video Fusion

Optimal vision is central to mission effectiveness of vehicle operations, situational awareness and reconnaissance, especially in degraded visual environments. Operation can be extended in adverse or low-light by using both daylight and infrared (IR) cameras. However, two displays can cause information overload and swamp data-links. RFEL's Digital Video Fusion solution overcomes this problem by producing an optimum composite image from both scenes. Now with new features,the user can have total command of scene definition across the visible and infrared spectra.  

Features

  • Significant performance advantage over blending and overlay methods
  • Image transformation and registration
  • Advanced multi-resolution combining with scalable rules
  • Contrast enhancement and optional filtering 
  • Low latency, optimised resource usage

New Features

  • Pre-processing input channel enhancement 
  • Colour noise suppression
  • Real-time selectable pseudo-colour depiction

Applications
Digital Video Fusion provides a sophisticated and effective solution to the problem of operating in diverse degraded visual environments. It is ideal for incorporation by product and system developers of; driver or pilot aids, surveillance, situational awareness and targeting system, and to enhance reconnaissance equipment. Designed typically to process images from daylight and infrared cameras, Fusion could equally well be used with other wideband, or even synthetic imagery. 

RFEL's Digital Video Fusion system assembles several high performance image processing techniques into a deliverable size, weight and power (SWaP) compatible solution. The system operates in real-time using FPGA processing augmented by embedded software. It is supported by a comprehensive suite of development tools, including reference designs and scripting for simulation, build and test that will accelerate the design of cutting-edge, multi-band vision system. The Fusion architecture exploits hybrid FPGA-ARM processors, like Xilinx Zynq or Altera SoC, dedicating FPGA processing elements to real-time functions, whilst leveraging the flexibility of a closely integrated configuration and control software host. 

  •   
  •   
  •   
  •   

Please enter the word you see in the image below:


 

Digital Fusion Diagram

Parameter Value
Image Registration Six dimensional, incl. perspective, user programable warping coefficients, deduced from software tool
Fusion Algorithm Adavnced multi-resolution fusion (not low performance blending) with real-time selectable fusion rules
Colour Processing Full colour fusion with graded reduction to monochrome that mimics the human eye transition from photopic (daylight) to scotpoic (low-light) vision
Image Contrast Enhancement Image contrast enhancement, incl. with histogram equalisation
Pre-processing  Propietary enhancement, and optional spatial or temporal filters. Sensor dependent
Feature Identification

Real-time selectable pseudo-colur: colour maps objects hottert than a specified threshold. Custom maps supported.

Videp i/p AXI-4 streaming, or frame memory mapped, interface as standard. Custom interfacing design service available
Video o/p AXI-4 frane memory mapped interface only as standard. Custom interfacing design service available
Pixel Depth IR: Up to 16-bit. DL: 16b YCbCr or 24b RGB. 
Resolution All standard formats upto 1080p
Frame Rate All standard formats up to 60fps
Host Interface AXI-4 Lite control interface. Custom interfacing design service available
Noise Suppression Optional spatial and/or temporal filtering
Total Latecy 2-3 frames; <50ms at 60fps (ultra low latency on request)
External Synchronisation Gen-lock & sensor sync. is sensor/application dependent; contact RFEL sales for advice on specific applications

Implementation
Depending on configuration, Digital Fusion requires a frame buffer memory; all interfaces (video, memory and control) are implemented as flexible AXI4 compliant interfaces or similar. As for all RFEL IP cores, Fusion is backed up by a detailed bit-true Matlab model of the processing. Contact RFEL for options to evaluate performance with suitable sample video sequences.

Warranty
For peace of mind RFEL provides a twelve month warranty as standard. If required, this can be extended as part of a contract at a reasonable extra cost.

The Dual Camera Demonstrator (DCD) has been developed to showcase RFEL's high performance dual-band, high-definition Fusion IP core. DCD demonstrates the effectiveness of HALO solutions in developing next generation vision systems, or to facilitate the deployment of improved capability on existing systems.

DCD was designed using HALO's Rapid Product Development System suite of qualified hardware and firmware IP modules. Designed as a platform for RFEL's complete suite of high-speed, real-time processing video processing IP functions, a HALO platform can be integrated with a user's system video and data interfaces, and can also seamlessly support user's design IP, at maximum performance.

Features

  • Integrated or external LWIR and Daylight camera optionsDual Camera Demonstrator Image
  • Flexible and easy to use
  • Fast RFEL IP evaluation and development system
  • Convenient off-the-shelf form-factor
  • Resourced to handle large applications

Applications

  • Defence surveillance, situational awareness, simulation and remote systems
  • Industrial and homeland security, perimeter protection and remote sensor inspection
  • Camera and sensor package development

The DCD is a standalone product suitable for desktop use, field and concept demonstrations. Supplied in a heavy duty enclosed chassis and ready to be operated out-of-the-box, it can be configured with commercial grade connectors to video and host interfaces, enabling rapid and convenient bring-up and adaption to user interfaces.

Latest News