EdgeMappingResearchProject/main.m at main · MGross21/EdgeMappingResearchProject · 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
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
% MICHAEL GROSS
% FALL 2022
% MAT 499

%% Setup Workspace

clear,clc
format compact
path_init();

% exist RESTOREDEFAULTPATH_EXECUTED var
% if(~ans)
%     restoredefaultpath
% end
% clc,clear ans

%% Setup Struct for Matrix Operators

mtrxOp.label = ["Prewitt", "Sobel", "Roberts Cross"];
mtrxOp.x = {([1,0,-1; 1,0,-1; 1,0,-1])
            ([-1,0,1;-2,0,2;-1,0,1])
            ([1,0;0,-1])};
mtrxOp.x = mtrxOp.x.';

mtrxOp.y = {([1,1,1; 0,0,0 ; -1,-1,-1])
            ([1,2,1;0,0,0;-1,-2,-1])
            ([0,-1;1,0])};
mtrxOp.y = mtrxOp.y.';

%% Apply Operators

%Initialize struct for image data
[iM.Count, iM.label, iM.path] = imLoad('EM Images', 'name');

% Create struct for holding results
iM.BW=[];
iM.RGB =[];

results=[];
results.raw={};
% Edgemapping Function for each image and operator combination
for n = 1: iM.Count % number of images
    data.temp={};
    for k = 1:size(mtrxOp.label,2) % number of matrix operators
        R = eMap(iM.label{n}, iM.path{n},mtrxOp.label{k},mtrxOp.x{k}, mtrxOp.y{k}, true, [227 227]);
        data.temp = [data.temp; R]; % storing data for each image
        results.raw = [results.raw; R]; %storing all data
    end
    results.(genvarname(iM.label{n})) = data.temp;
    iM.RGB = [iM.RGB, data.temp(1,1)];
    iM.BW = [iM.BW, data.temp(1,6)];

end
clear("k","n","R","data")

%% Object Detection Training

%Run training data if it doesn't already exist
oD.parentPath = "D:\Michael Gross\Documents\MAT499 Images\";
oD.oldPath = "Raw Images";
oD.newPath = "Training Data";
oD.catNames = ["bicycle","car", "pedestrian","street sign","trailer truck"];

oD.pAdd = oD.parentPath + [oD.oldPath; oD.newPath] + "\" + oD.catNames;
for i = 1:numel(oD.pAdd)
    addpath(oD.pAdd{i})
end

if(~(exist("myNet.mat", "file")>0))
    [oD.catCount] = imLoad(oD.parentPath + oD.oldPath, 'folder');

    for i = 1:numel(oD.catNames)
        [oD.parentPath,oD.newPath] = trimData(oD.parentPath, oD.oldPath ,oD.catNames(i), oD.newPath, 5000);
    end

    [oD.myNet, oD.testiM]  = transferLearn(oD.parentPath + oD.newPath, oD.catCount);

    myNet = oD.myNet;
    save("myNet.mat", "myNet");
else
    oD.myNet = load("myNet.mat","myNet");
end

clear i
%% Test/Record Confidence Rating

% oD.store = imageDatastore("EMaps\",'IncludeSubfolders', true, 'LabelSource', 'foldernames');
% oD.labCount = countEachLabel(oD.store);
% for c = 1:oD.catCount
%     oD.split{c} = splitEachLabel(oD.store,sum(oD.store.Labels==oD.catNames(c)),"Include",oD.catNames(c));
% end
idx = 1;
for i = 1:numel(oD.catNames)
    if i==1
        oD.conRat = [];
        oD.reSize = [];
    end
    oD.eMiM = [results.(genvarname(iM.label{i}))(1,6); results.(genvarname(iM.label{i}))(1:numel(mtrxOp.x),9)]; %BW + Gradient Images

    for j = 1:numel(oD.eMiM)
        imwrite(oD.eMiM{j},strcat("ExportedPictures\EMaps\",iM.label(i),string(j),".png"))

        [temp_predictedLabel, temp_confidence, temp_accuracy]  = confidenceRating(@readFunctionTrain, oD.myNet.myNet, readFunctionTrain("",oD.eMiM{j}), oD.catNames(i));
        results.(genvarname(iM.label{i}))(j,11:13) = [{string(temp_predictedLabel)},{max(temp_confidence)},{temp_accuracy}];
        if j > 1
            results.raw(idx,11:13) = [{string(temp_predictedLabel)},{max(temp_confidence)},{temp_accuracy}];
            idx = idx+1;
        end
    end

%     oD.reSize = readFunctionTrain("",oD.eMiM);
    clear temp_predictedLabel temp_confidence temp_accuracy
end
clear i c j

%% Plot subplot results
% passes in results for each image
for m = 1:iM.Count
    sPlot(iM.label{m}, iM.RGB{m}, iM.BW{m}, mtrxOp.label, results.(genvarname(iM.label{m}))); %passes in image (label,RGB, BW), operator labels, and results
    pause
end
clear m

%% Save data

results.header = ["Original Image", "Width","Height","Total Size","Operator","B/W Image","Horizontal Gradient","Vertical Gradient","EMap Gradient","Time","Classification","Confidence Rating","Correct"];

save finalResults.mat
%images cannot export to excel, save instead

% for m = 1:iM.Count
%     locations = [{[1,1]},{[1,9]},{[2,9]},{[3,9]}]; %BW + Grad Image result Locations
%
%     imDownload(results.(genvarname(iM.label{m})), iM.label{m},locations, "ExportedImages")
% end

xPort("EdgeMapResults.xlsx", results.raw, results.header, [2:5,10:13])