GPU Code Generation for a Fog Rectification Simulink Model
This example demonstrates how to generate CUDA® code from the Simulink® model that takes a foggy image as input and produces a defogged image as output. This example is a typical implementation of fog rectification algorithm. The example uses conv2, rgb2gray, and imhist (Image Processing Toolbox) functions. This example closely follows Fog Rectification example. This example illustrates the following concepts:
Verification of GPU Environment.
Model fog rectification application in Simulink by using image processing functions.
Configure the model for GPU code generation.
Generate a CUDA executable for the Simulink model.
Prerequisites
CUDA enabled NVIDIA® GPU with compute capability 3.2 or higher.
NVIDIA CUDA toolkit and driver.
Image Processing Toolbox™.
Environment variables for the compilers and libraries. For more information, see Third-Party Hardware and Setting Up the Prerequisite Products.
Verify GPU Environment
To verify that the compilers and libraries necessary for running this example are set up correctly, use the coder.checkGpuInstall function.
envCfg = coder.gpuEnvConfig('host');
envCfg.BasicCodegen = 1;
envCfg.Quiet = 1;
coder.checkGpuInstall(envCfg);
Fog Rectification Simulink Model
The Simulink model for fog rectification consists of Fog Rectification subsystem that contains a MATLAB Function block which takes a foggy image as input and returns a defogged image as output. It uses fog_rectification algorithm described in Fog Rectification example. When the model runs, the Visualization block displays the foggy input image and defogged output image.
open_system('fog_rectification_model');
Run model on the Simulation Target
Open Configuration Parameters dialog box.
In Simulation Target pane, select GPU acceleration.
Run the Simulation in Normal mode.
set_param('fog_rectification_model', 'SimulationMode', 'Normal'); sim('fog_rectification_model');
Generate and Build the Simulink Model
In Code Generation pane, select the Language as C++ and enable Generate GPU code.
Open Code Generation > GPU Code pane. In the subcategory Libraries, enable cuBLAS, cuSOLVER and cuFFT.
Generate and build the Simulink model on the host GPU by using the rtwbuild command. The code generator places the files in a build folder, a subfolder named fog_rectification_model_ert_rtw under your current working folder.
status = evalc("rtwbuild('fog_rectification_model')");
Cleanup
Close the Simulink model.
close_system('fog_rectification_model');