Vision HDL Toolbox
Design image processing, video, and computer vision systems for FPGAs and ASICs
Vision HDL Toolbox™ provides pixel-streaming algorithms for the design and implementation of vision systems on FPGAs and ASICs. It provides a design framework that supports a diverse set of interface types, frame sizes, and frame rates. The image processing, video, and computer vision algorithms in the toolbox use an architecture appropriate for HDL implementations.
The toolbox algorithms are designed to generate readable, synthesizable code in VHDL® and Verilog® (with HDL Coder™). The generated HDL code is FPGA-proven for frame sizes up to 8k resolution and for high frame rate (HFR) video.
Toolbox capabilities are available as MATLAB® functions, System objects™, and Simulink® blocks.
Start building your automated driving system with hardware-proven subsystems for lane detection, pothole detection, and stereo disparity computation.
Learn how to implement feature detection techniques with streaming hardware to develop surveillance, object tracking, industrial inspection, and other applications.
Jumpstart development of image conditioning hardware using examples of noise removal, gamma correction, and histogram implementations.
Hardware-Accelerated Vision Processing
Model and simulate efficient hardware implementations of vision processing algorithms, such as conversions, filtering, morphology, and statistics. Then use HDL Coder to generate synthesizable VHDL or Verilog RTL.
Processing Multiple Pixels Per Clock
Process 4k, 8k, or high-frame-rate video at FPGA clock rates by specifying parallel streams of 4 or 8 pixels. The underlying hardware implementation is automatically updated to support simulation and code generation with the specified parallelism.
Built-In Hardware Data Management
Use Vision HDL Toolbox blocks to automatically manage streaming input data, such as control signals, region-of-interest (ROI) windows, and line buffers. Use HDL Coder to generate VHDL or Verilog RTL for the control functionality you model and simulate.
Conversion Between Frames and Pixels
Convert full-frame video to a stream of pixels with control signals for processing in hardware. Then convert the streaming hardware output to frames for verification against your golden reference algorithm.
Prototype Platform with Live Video Input
Prototype your vision processing application by downloading the Computer Vision Toolbox Support Package for Xilinx® Zynq®-Based Hardware and using HDL Coder and Embedded Coder® to generate code from your MATLAB or Simulink implementation.
Use HDL Coder to generate high-quality, target-independent RTL and AXI interfaces from your hardware subsystem models.
Vision Processing for FPGA
Watch this five-part video series that introduces key concepts and the workflow for targeting vision applications to FPGAs for prototyping and production.
Corner Detector Block and System Object
Detect features using FAST algorithm
Line Buffer with No Padding
Specify option to not add padding for blocks that use line buffer memory
Down-size an image frame by a specified factor
External Memory Modeling Examples
Learn how to model vision algorithms that require frame buffering in external memory (Requires SoC Blockset)
Computer Vision on Xilinx Zynq-Based Hardware
Target a Zynq UltraScale+ MPSoC with an Avnet FMC-HDMI-CAM module
Fog Rectification Example
Enhance hazy images to improve clarity