Description

Abstract

Background subtraction (BGS) is a commonly used technique for achieving this segmentation. Background subtraction is a widely used approach to detect moving objects from static and dynamic cameras. Many different methods have been proposed over the recent years and there are a number of object extraction algorithms proposed in this survey it has most efficiently constrained environments where the background is relatively easy and static. In this paper, we analysis most popular, state-of- the-art BGS algorithms and propose a neuro fuzzy model for determining thresholds, we examine how threshold algorithm poor their performance. Our method shows that threshold plays a major role in obtaining the foreground segmentation masks produced by a BGS algorithm and our experimental results demonstrate that neuro fuzzy system is much more accuracy and robust than existing system approaches.

Demonstration Video

Matlab code for Background subtraction using Fuzzy

clc;

clear;
close all;
a=imread(‘b.bmp’);
f=imread(‘1.bmp’);
g=imread(‘g1.bmp’);
g=im2bw(g);
a=rgb2gray(a);
f=rgb2gray(f);

% K1=back_final(a,f);
K1=imsubtract(a,f);

K11=K1(:);
K11=double(K11);
fis = readfis(‘binary’);
Out = evalfis([K11],fis)

Out=round(Out);
Out=reshape(Out,[128 160]);
figure;
imshow(Out,[]);
Out=medfilt2(Out,[5 5]);

Out=imcomplement(Out);
%%%%%%%%%%%%%%%%%%%%%phase 2
[L num]=bwlabel(Out);

STATS=regionprops(L,’all’);
cc=[];
removed=0;
% h = waitbar(0,’Please wait…’);
for i=1:num
dd=STATS(i).Area;
if (dd < 1000)

L(L==i)=0;
removed = removed + 1;
num=num-1;
% imshow(L==i);
else
% imshow(L==i);
end

% waitbar(i/num,h)
end

[L2 num2]=bwlabel(L);

subplot(2,2,1);imshow(a);
subplot(2,2,2);imshow(f);
subplot(2,2,3);imshow(L2);
subplot(2,2,4);imshow(g);

[Tp,Tn,Fp,Fn,Sensitivity,Specificity]= performanceanalysis(g,L2)
disp(Sensitivity);
disp(Specificity);
disp(‘exit’);

Results