High Dynamic Range Imaging

This example shows how to generate HDL code from a MATLAB® design that implements a high dynamic range imaging algorithm.

Algorithm

High Dynamic Range Imaging (HDRI or HDR) is a set of methods used in imaging and photography to allow a greater dynamic range between the lightest and darkest areas of an image than current standard digital imaging methods or photographic methods. HDR images can represent more accurately the range of intensity levels found in real scenes, from direct sunlight to faint starlight, and is often captured by way of a plurality of differently exposed pictures of the same subject matter.

MATLAB Design

design_name = 'mlhdlc_hdr';
testbench_name = 'mlhdlc_hdr_tb';

Let us take a look at the MATLAB design

dbtype(design_name);
1     function [valid_out, x_out, y_out, ...
2         HDR1, HDR2, HDR3] = mlhdlc_hdr(YShort1, YShort2, YShort3, ...
3         YLong1, YLong2, YLong3, ...
4         plot_y_short_in, plot_y_long_in, ...
5         valid_in, x, y)
6     %
7
8     %   Copyright 2013-2015 The MathWorks, Inc.
9
10    % This design implements a high dynamic range imaging algorithm.
11
12    plot_y_short = plot_y_short_in;
13    plot_y_long = plot_y_long_in;
14
15    %% Apply Lum(Y) channels LUTs
16    y_short = plot_y_short(uint8(YShort1)+1);
17    y_long = plot_y_long(uint8(YLong1)+1);
18
19    y_HDR = (y_short+y_long);
20
21    %% Create HDR Chorm channels
22    % HDR per color
23
24    HDR1 = y_HDR * 2^-8;
25    HDR2 = (YShort2+YLong2) * 2^-1;
26    HDR3 = (YShort3+YLong3) * 2^-1;
27
28    %% Pass on valid signal and pixel location
29
30    valid_out = valid_in;
31    x_out = x;
32    y_out = y;
33
34    end
dbtype(testbench_name);
1
2     %
3
4     %   Copyright 2013-2015 The MathWorks, Inc.
5
6     % Clean screen and memory
7     close all
8     clear mlhdlc_hdr
9     set(0,'DefaultFigureWindowStyle','docked')
10
11
12    %% Read the two exposed images
13
16
17    % define HDR output variable
18    HDR = zeros(size(short));
19    [height, width, color] = size(HDR);
20
21    figure('Name', [mfilename, '_plot']);
22    subplot(1,3,1);
23    imshow(short, 'InitialMagnification','fit'), title('short');
24
25    subplot(1,3,2);
26    imshow(long, 'InitialMagnification','fit'), title('long');
27
28
29    %% Create the Lum(Y) channels LUTs
30    % Pre-process
31    % Luminance short LUT
32    ShortLut.x = [0    16    45    96   255];
33    ShortLut.y = [0    20    38    58   115];
34
35    % Luminance long LUT
36    LongLut.x = [ 0 255];
37    LongLut.y = [ 0  140];
38
39    % Take the same points to plot the joined Lum LUT
40    plot_x = 0:1:255;
41    plot_y_short = interp1(ShortLut.x,ShortLut.y,plot_x); %LUT short
42    plot_y_long = interp1(LongLut.x,LongLut.y,plot_x); %LUT long
43
44    %subplot(4,1,3);
45    %plot(plot_x, plot_y_short, plot_x, plot_y_long, plot_x, (plot_y_long+plot_y_short)), grid on;
46
47
48    %% Create the HDR Lum channel
49    % The HDR algorithm
50    % read the Y channels
51
52    YIQ_short = rgb2ntsc(short);
53    YIQ_long = rgb2ntsc(long);
54
55    %% Stream image through HDR algorithm
56
57    for x=1:width
58        for y=1:height
59            YShort1 = round(YIQ_short(y,x,1)*255); %input short
60            YLong1 = round(YIQ_long(y,x,1)*255); %input long
61
62            YShort2 = YIQ_short(y,x,2); %input short
63            YLong2 = YIQ_long(y,x,2); %input long
64
65            YShort3 = YIQ_short(y,x,3); %input short
66            YLong3 = YIQ_long(y,x,3); %input long
67
68            valid_in = 1;
69
70            [valid_out, x_out, y_out, HDR1, HDR2, HDR3] = mlhdlc_hdr(YShort1, YShort2, YShort3, YLong1, YLong2, YLong3, plot_y_short, plot_y_long, valid_in, x, y);
71
72            % use x and y to reconstruct image
73            if valid_out == 1
74                HDR(y_out,x_out,1) = HDR1;
75                HDR(y_out,x_out,2) = HDR2;
76                HDR(y_out,x_out,3) = HDR3;
77            end
78        end
79    end
80
81    %% plot HDR
82    HDR_rgb = ntsc2rgb(HDR);
83    subplot(1,3,3);
84    imshow(HDR_rgb, 'InitialMagnification','fit'), title('hdr ');

Simulate the Design

It is always a good practice to simulate the design with the testbench prior to code generation to make sure there are no runtime errors.

mlhdlc_hdr_tb Setup for the Example

Executing the following lines copies the necessary files into a temporary folder

mlhdlc_demo_dir = fullfile(matlabroot, 'toolbox', 'hdlcoder', 'hdlcoderdemos', 'matlabhdlcoderdemos');
mlhdlc_temp_dir = [tempdir 'mlhdlc_hdr'];

% create a temporary folder and copy the MATLAB files
cd(tempdir);
[~, ~, ~] = rmdir(mlhdlc_temp_dir, 's');
mkdir(mlhdlc_temp_dir);
cd(mlhdlc_temp_dir);

% copy files to the temp dir
copyfile(fullfile(mlhdlc_demo_dir, [design_name,'.m*']), mlhdlc_temp_dir);
copyfile(fullfile(mlhdlc_demo_dir, [testbench_name,'.m*']), mlhdlc_temp_dir);
copyfile(fullfile(mlhdlc_demo_dir, 'mlhdlc_hdr_long.tif'), mlhdlc_temp_dir);
copyfile(fullfile(mlhdlc_demo_dir, 'mlhdlc_hdr_short.tif'), mlhdlc_temp_dir);

Create a New HDL Coder™ Project

coder -hdlcoder -new mlhdlc_hdr_prj

Next, add the file 'mlhdlc_hdr.m' to the project as the MATLAB Function and 'mlhdlc_hdr_tb.m' as the MATLAB Test Bench.

Refer to Get Started with MATLAB to HDL Workflow for a more complete tutorial on creating and populating MATLAB HDL Coder projects.

Creating constant parameter inputs

This example shows to use pass constant parameter inputs.

In this design the input parameters 'plot_y_short_in' and 'plot_y_long_in' are constant input parameters. You can define them accordingly by modifying the input types as 'constant(double(1x256))'

'plot_y_short_in' and 'plot_y_short_in' are LUT inputs. They are constant folded as double inputs to the design. You will not see port declarations for these two input parameters in the generated HDL code.

Note that inside the design 'mlhdlc_hdr.m' these variables are reassigned so that they get properly fixed-point converted. This is not necessary if these are purely used as constants for defining sizes of variables for example and not part of the logic.

Run Fixed-Point Conversion and HDL Code Generation

Launch HDL Advisor and right click on the 'Code Generation' step and choose the option 'Run to selected task' to run all the steps from the beginning through the HDL code generation.

Convert the design to fixed-point and generate HDL code

The following script converts the design to fixed-point, and generate HDL code with a test bench.

exArgs = {0,0,0,0,0,0,coder.Constant(ones(1,256)),coder.Constant(ones(1,256)),0,0,0};
fc = coder.config('fixpt');
fc.TestBenchName = 'mlhdlc_hdr_tb';
hc = coder.config('hdl');
hc.GenerateHDLTestBench = true;
hc.SimulationIterationLimit = 1000; % Limit number of testbench points
codegen -float2fixed fc -config hc -args exArgs mlhdlc_hdr

Examine the generated HDL code by clicking on the hyperlinks in the Code Generation Log window.

Clean up the Generated Files

You can run the following commands to clean up the temporary project folder.

mlhdlc_demo_dir = fullfile(matlabroot, 'toolbox', 'hdlcoder', 'hdlcoderdemos', 'matlabhdlcoderdemos');
mlhdlc_temp_dir = [tempdir 'mlhdlc_hdr'];
clear mex;
cd (mlhdlc_demo_dir);
rmdir(mlhdlc_temp_dir, 's');