Computer_Vision--Background-Subtraction-Algorithm-with-Single-Gaussian-Model/Computer_Vision_Background_Subtraction.m at main · VimsRocz/Computer_Vision--Background-Subtraction-Algorithm-with-Single-Gaussian-Model · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71

% Authors: <Chawda Vimal>

close all
clear all

%% Read image sequence
path2sequence = 'sequence';
search_string = fullfile(path2sequence, '*.jpeg');
file_list = dir(search_string);

%% Sequencial estimation of the gaussians parameters

% TODO: Learning rate
% alpha = ?;

% Initial values
im_RGB = imread(fullfile(path2sequence, file_list(1).name));
[m,n] = size(rgb2gray(im_RGB));
n1=50;
n2 = 1400;
alpha = 1/n2;
mu = single(rgb2gray(im_RGB));
sigma_square = ones(m,n)*100;

% Structure element for morphological operation
se = strel('square',3);

h1 = figure(1);
colormap(gray);

% Iterate over the sequence
for i = 2:length(file_list)
    colormap(gray);
    im_RGB = imread(fullfile(path2sequence, file_list(i).name));
    im = single(rgb2gray(im_RGB));

    % TODO: Thresholding for background subtraction
    %delta_g = ?
    %return_mask = ? -> background pixels are 1, foreground pixels are set to 0
    %return_mask = logical(return_mask); %convert to type "logical"
     del_g =abs(im-mu);                                                                                                                                               variable =2.5*sqrt(sigma_square);
     B=del_g<variable;
     return_mask = logical(B);
     before_mask =return_mask;

    % Eliminate noise
    % TODO: Closing using structure element "se"
    return_mask = imdilate(return_mask,se);
    return_mask = imerode(return_mask,se);

    % Update Gaussian parameters
    % TODO: Mean
    % TODO: Variance
    mu = (1-alpha)*mu+alpha*im;
    sigma_square = (1-alpha)*sigma_square + alpha*((mu-im).^2);
    % Output
    subplot(231);
    imagesc(im_RGB, [0 255]); % TODO: Show input image   we can write or im or im_RGB
    subplot(232);
    imagesc(mu, [0 255]); % TODO: Show mean values
    subplot(233);
    imagesc(sigma_square, [0 255]); % TODO: Show variances
    subplot(234);
    imagesc(del_g, [0 255]); % TODO: Show difference between current image and mean values
    subplot(235);
    imagesc(before_mask, [0 1]); % TODO: Show background mask (unfiltered)
    subplot(236);
    imagesc(return_mask, [0 1]); % TODO: Show background mask after eliminating noise
    drawnow;
end