DSAR-CF/run_DSARCF.m at master · RuizeHan/DSAR-CF · 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
function results = run_DSARCF(seq, ~, ~, ~)

% By RuizeHan
% Dynamic Saliency-Aware Regularization for Correlation Filter based Object Tracking
% In TIP 2019

close all
addpath(genpath('Processing/'));

seq.format = 'otb';
% set the parameters of the method
params = SetParams(seq);

% prepare the data of the method
[params, data] = PrepareData(params);

time = 0;
timeF = 0;
timeD = 0;

for frame = 1:data.seq.num_frames
    data.seq.frame = frame;
    data.seq.im = imread(params.s_frames{data.seq.frame});
    if size(data.seq.im,3) > 1 && data.seq.colorImage == false
        data.seq.im = data.seq.im(:,:,1);
    end

    tic();
    % Detection process of CF tracking
    [params, data] = Detection(params, data);
    timeD = timeD + toc();

    time = toc;

    if rem(frame+params.update-1,params.update) == 0
    % initial the correlation filter
    if params.selector == -1 && params.selector_last == 1
        data.conf = getFilterSetup(params,data,'conf');
    end

    % update the correlation filter
    [params, data] = FilterUpdate(params, data);
    end
    timeF = timeF + toc() -time;
    timeW = data.time.timeW;
    timeS = data.time.timeS;
    time = timeD + timeF;

    Visualization(params.visualization, params.selector, data.seq.frame, data.seq.im, data.obj.pos, data.obj.target_sz);

end

% get the running fps of DSARCF
fps = numel(params.s_frames) / time;

% compute the running time of each component
% disp(['fps: ' num2str(fps)])
results.time.timeF = timeF;
results.time.timeD = timeD;
results.time.time = time;
results.time.timeW = timeW;
results.time.timeS = timeS;
results.type = 'rect';
results.res = data.obj.rects; % each row is a rectangle
results.fps = fps;

end