Scribd
Upload
120 views7 pages

Tone Mapping

Tone Mapping is the process to convert HDR(High Dynamic Range) image into LDR(Low Dynamic Range) image.

Uploaded by

Mahvish Fatima
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
120 views7 pages

Tone Mapping

Tone Mapping is the process to convert HDR(High Dynamic Range) image into LDR(Low Dynamic Range) image.

Uploaded by

Mahvish Fatima
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 7

HDR Images at different Exposures:

1. Size (210 KB (216,011 bytes)

2. Size 181 KB (186,083 bytes)

3. Size 156 KB (160,250 bytes)


Tone Mapping

Reinchard Tone mapping

m.file
function result = apply_reinhard_global_tonemap(hdr_map, a)
fprintf('== Applying Reinhard Global Tonemap (a = %.3f) ==\n', a);
[height, width, num_channels] = size(hdr_map);
num_pixels = height * width;
delta = 0.0001;
luminance_map = compute_luminance_map(hdr_map);
C = double(delta + luminance_map);
key = exp((1 / num_pixels) * sum(sum(log(C))));
scaled_luminance = (a / key) * luminance_map;
display_luminance = scaled_luminance ./ (1 + scaled_luminance);
result = apply_tonemap(hdr_map, luminance_map, display_luminance);
end

Tone Mapping Result:

After Tone Mapping..Size become as 111.799 KB


JPEG:

Now applying JPEG on this tone map output

M.file
function Anew=imgreader(x)
%reading the image
A=imread(x);

%get the number of pixels


[rows,columns] = size(A);
disp('No of pixels')
Number_Pixels = rows*columns;

%convert it to grayscale
%AS=rgb2gray(A);
AS=A;

%get the row and col size


rowSize=size(AS,1);
colSize=size(AS,2);

%subtract the bytes from the image


s=int16(AS)-128;
B=[];
Bq=[];
count=1;

blockSize=input('Enter 8 to encode using 8X8 or 16 to use 16X16 : ')


jump=0;
zzcount=0;
if blockSize==8
jump=7;
zzcount=64;
printLimit=8;
else
jump=15;
zzcount=256;
printLimit=16;
end

%encoding process
for i=1:blockSize:rowSize
for j=1:blockSize:colSize
%performing the DCT
B(i:i+jump,j:j+jump) = dct2(s(i:i+jump,j:j+jump));
%performing the quantization
Bq(i:i+jump,j:j+jump)=quantization(B(i:i+jump,j:j+jump),blockSize);
zigzag(count,1:zzcount)=zzag(Bq(i:i+jump,j:j+jump));
count=count+1;
end
end
disp('Original Image')
AS(1:printLimit,1:printLimit)
disp('After level shift')
s(1:printLimit,1:printLimit)
disp('After DCT')
B(1:printLimit,1:printLimit)
disp('After Quantization')
Bq(1:printLimit,1:printLimit)
zigzag(1,1:zzcount);
%for putting it into a file and compressing
save('compImg.mat','zigzag');
compImage=gzip('compImg.mat')

%for decompressing the file


gunzip('compImg.mat.gz')
%fid = fopen('magic5.bin', 'r');
%Bq=load('compImg.mat')

Bnew=[];
ASnew=[];
%decoding process
for i=1:blockSize:rowSize
for j=1:blockSize:colSize
%performing the dequantization

Bnew(i:i+jump,j:j+jump)=invQuantization(Bq(i:i+jump,j:j+jump),blockSize);
%performing the inverse DCT
ASnew(i:i+jump,j:j+jump) = round(idct2(Bnew(i:i+jump,j:j+jump)));
end
end
disp('dequantization')
%Bnew(1:printLimit,1:printLimit)
disp('Inverse DCT')
%ASnew(1:printLimit,1:printLimit)
Anew=ASnew+128;
Anew=uint8(Anew);
rgbImage = cat(3,Anew,Anew,Anew);

figure
imshow(A)
title('Original Image')
figure
imshow(Anew)
title('Compressed Image')

im1_stats = dir('new.jpg');
aw=im1_stats;
p=aw.bytes;
msgbox({sprintf('Original Image,%g (KB)',p/1000)});
im1_stats = dir('Compressed Image');
aw=im1_stats;
pp1=aw.bytes;
msgbox({sprintf('Compressed Image,%g (KB)',pp1/1000)});
end
Might be wrong…but achieve Compressed Image size become as 29.173KB

You might also like