用matlab 对图像进行算术编码和解码的程序!!!求大神指点
发布网友
发布时间：2022-04-26 20:07
共1个回答
热心网友
时间：2023-10-27 09:55
close all; clear all; clc; %关闭所有图形窗口，清除工作空间所有变量，清空命令行
I=imread('lena.bmp');%待编码的矩阵
imshow(I);
thresh = graythresh(I); %自动确定二值化阈值
I2 = im2bw(I,thresh); %对图像二值化
imshow(I2);
[m,n]=size(I2); %计算矩阵大小
I2=double(I2);
p_table=tabulate(I2(:)); %统计矩阵中元素出现的概率，第一列为矩阵元素，第二列为个数，第三列为概率百分数
color=p_table(:,1)';
p=p_table(:,3)'/100; %转换成小数表示的概率
psum=cumsum(p_table(:,3)'); %计算数组各行的累加值
allLow=[0,psum(1:end-1)/100];%由于矩阵中元素只有两种，将[0,1）区间划分为两个区域allLow和 allHigh
allHigh=psum/100;
numberlow=0; %定义算术编码的上下限numberlow和numberhigh
numberhigh=1;
for k=1:m %以下计算算术编码的上下限，即编码结果
for kk=1:n
data=I2(k,kk);
low=allLow(data==color);
high=allHigh(data==color);
range=numberhigh-numberlow;
tmp=numberlow;
numberlow=tmp+range*low;
numberhigh=tmp+range*high;
end
end
fprintf('算术编码范围下限为%16.15f\n\n',numberlow);
fprintf('算术编码范围上限为%16.15f\n\n',numberhigh);
Mat=zeros(m,n); %解码
for k=1:m
for kk=1:n
temp=numberlow<low;
temp=[temp 1];
indiff=diff(temp);
indiff=logical(indiff);
Mat(k,kk)=color(indiff);
low=low(indiff);
high=allHigh(indiff);
range=high - low;
numberlow=numberlow-low;
numberlow=numberlow/range;
end
end
热心网友
时间：2023-10-27 09:55
close all; clear all; clc; %关闭所有图形窗口，清除工作空间所有变量，清空命令行
I=imread('lena.bmp');%待编码的矩阵
imshow(I);
thresh = graythresh(I); %自动确定二值化阈值
I2 = im2bw(I,thresh); %对图像二值化
imshow(I2);
[m,n]=size(I2); %计算矩阵大小
I2=double(I2);
p_table=tabulate(I2(:)); %统计矩阵中元素出现的概率，第一列为矩阵元素，第二列为个数，第三列为概率百分数
color=p_table(:,1)';
p=p_table(:,3)'/100; %转换成小数表示的概率
psum=cumsum(p_table(:,3)'); %计算数组各行的累加值
allLow=[0,psum(1:end-1)/100];%由于矩阵中元素只有两种，将[0,1）区间划分为两个区域allLow和 allHigh
allHigh=psum/100;
numberlow=0; %定义算术编码的上下限numberlow和numberhigh
numberhigh=1;
for k=1:m %以下计算算术编码的上下限，即编码结果
for kk=1:n
data=I2(k,kk);
low=allLow(data==color);
high=allHigh(data==color);
range=numberhigh-numberlow;
tmp=numberlow;
numberlow=tmp+range*low;
numberhigh=tmp+range*high;
end
end
fprintf('算术编码范围下限为%16.15f\n\n',numberlow);
fprintf('算术编码范围上限为%16.15f\n\n',numberhigh);
Mat=zeros(m,n); %解码
for k=1:m
for kk=1:n
temp=numberlow<low;
temp=[temp 1];
indiff=diff(temp);
indiff=logical(indiff);
Mat(k,kk)=color(indiff);
low=low(indiff);
high=allHigh(indiff);
range=high - low;
numberlow=numberlow-low;
numberlow=numberlow/range;
end
end
