ActivityTrackingWithSensorFusion/GroundTruth.py at last-working · harrywaugh/ActivityTrackingWithSensorFusion · 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
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
import numpy as np
import matplotlib
from matplotlib import pyplot as plt
from scipy import integrate
from scipy import interpolate
import math
from scipy import signal
import copy
import os

class GroundTruthGPX:
    latlngs = np.asmatrix([0.0, 0.0])
    def __init__(self, filename, gps, load_cache=False):
#         if(load_cache):
#             self.read_from_cache(filename)
#             return

#         directory = 'Data/ground_truth/' + filename + '.gpx'
#         ##Parse GT file
#         f = open(directory, 'r')
#         latlngs = []
#         #Read Data from CSV
#         for line in f:
#             tags = [tag[6:] for tag in line.split("><") if tag[0]=='r']
#             for tag in tags:
#                 split_tag = tag.split('"')
#                 if(len(split_tag)>1):
#                     latlngs.append([float(split_tag[3]), float(split_tag[1])])
#         self.latlngs= np.asmatrix(latlngs)
#         self.dis = self.convert_longlat_to_dis(self.latlngs)

#         ## Add  linear time to GPS, interpolated from start time to end time

#         time_vector = []
#         indexes = []
#         for position in self.dis:
#             distances = gps[:, 1:3] - position
#             magnitudes = np.linalg.norm(distances, ord = 2, axis = 1)
#             indexes.append(np.argmin(magnitudes))
#             time_vector.append([gps[indexes[-1], 0]])

#         self.dis = np.concatenate((np.asmatrix(time_vector), self.dis), axis=1)
#         self.dis[-1] = gps[-1, 0:3]
#         indexes[-1] = len(gps)

#         indexes.sort()
#         previous_index = indexes[0]
#         interpolated_positions = []
#         for i in range(1, len(indexes)):
#             n = indexes[i] - previous_index

#             if (n > 1):
#                 a = self.dis[i-1]
#                 b = self.dis[i]
#                 interpolated_positions.extend(self.interpolate(a, b, n))
#             elif(n == 1):
#                 interpolated_positions.append(self.dis[i])
#             previous_index = indexes[i]

#         self.dis = np.squeeze(np.asarray(interpolated_positions))
#         latlngs = np.asmatrix([0.0, 0.0])


#         print(len(gps))
#         print(len(self.dis))
        if(load_cache):
            self.read_from_cache(filename)
            return

        directory = 'Data/ground_truth/' + filename + '.gpx'
        ##Parse GT file
        f = open(directory, 'r')
        latlngs = []

        for line in f:
            tags = [tag[6:] for tag in line.split("><") if tag[0]=='r']
            for tag in tags:
                split_tag = tag.split('"')
                if(len(split_tag)>1):
                    latlngs.append([float(split_tag[3]), float(split_tag[1])])
        self.latlngs= np.asmatrix(latlngs)

        lin_time = np.asmatrix(np.linspace(0, gps[-1, 0], len(self.latlngs))).T
        self.latlngs = np.concatenate((lin_time, self.latlngs), axis=1)
        irreg_var = self.latlngs
        reg_varX = np.asmatrix(np.interp(gps[:, 0], np.ravel(irreg_var[:,0]), np.ravel(irreg_var[:,1])))
        reg_varY = np.asmatrix(np.interp(gps[:, 0], np.ravel(irreg_var[:,0]), np.ravel(irreg_var[:,2])))
        self.latlngs = np.concatenate((reg_varX, reg_varY), axis=1)


        self.dis = self.convert_longlat_to_dis(self.latlngs)
        self.dis = np.concatenate((gps[:, 0], self.dis), axis=1)

        self.correct_ground_truth(self.dis[:, 1:3], gps[:,1:3])
        print(len(self.dis))
        self.dis = np.concatenate((gps[:, 0], self.dis), axis=1)

        print("Aligned Ground Truth")

    def interpolate(self, a, b, n):
        interp_points = []
        for i in range (n):
            interp_points.append(a + ((b - a) * (float(i))) / (n-1))
        return interp_points

    def convert_longlat_to_dis(self, gps):
        dis_list = [[0.0, 0.0]]
        start_gps = gps[0]
        for i in range(1, gps.shape[0]):
            dis_list.append([self.get_arc_len(start_gps[0, 0], gps[i, 0]),
                             self.get_arc_len(start_gps[0, 1], gps[i, 1])])
        return np.asmatrix(dis_list)

    def get_arc_len(self, deg1, deg2):
        delta_theta = deg2 - deg1
        delta_theta = ((delta_theta+180)%360)-180
        earth_R = 6378100
        return 2.0*math.pi*earth_R*delta_theta/360.0

    def correct_ground_truth(self, ground_truth, gps):
        ## Function that identifies the closest ground truth point for a given sensor reading
        def closest_ground_truth_point(x):
            ## Check every 200 steps
#             check_step = 300
            ground_truth_steps = ground_truth
            ## Get a vector, that represents the distance to all ground truth points
            distance_vectors = ground_truth_steps - x
            ## Get the magnitude, and thus identify the index with the minimum distance
            magnitudes = np.linalg.norm(distance_vectors, ord = 2, axis = 1)
            min_step_index = np.argmin(magnitudes)


#             ## Check points around close point
#             lower_bound = max(0,                   int(min_step_index*check_step - (check_step/2)))
#             upper_bound = min(len(ground_truth)-1, int(min_step_index*check_step + (check_step/2)))
#             neighbours = ground_truth[lower_bound : upper_bound]

#             ##Create distance vector, get magintudes and get closest two points
#             distance_vectors = neighbours - x
#             magnitudes = np.linalg.norm(distance_vectors, ord = 2, axis = 1)
#             minimum_neighbour = np.argmin(magnitudes)

            return ground_truth_steps[min_step_index]

        for i in range(len(gps)):
            ground_truth[i] = closest_ground_truth_point(gps[i])
        self.dis = ground_truth

    def read_from_cache(self, filename):
        self.dis = self.read_var_from_cache(filename, "dis")
#         self.latlngs = self.read_var_from_cache(filename, "latlng")
        print("READ GROUND TRUTH DATA FROM CACHE")

    def read_var_from_cache(self, filename, var_name):
        f=open('Data/ground_truth/cache/'+var_name+'/'+filename+'.csv',"r")
        var=[]
        for line in f:
            split = line.split(',')
            var.append([float(split[0]), float(split[1]), float(split[2])])
        f.close()
        return np.asmatrix(var)

    def write_to_cache(self, filename):
        self.write_var_to_cache(filename, "dis", self.dis)
#         self.write_var_to_cache(filename, "latlng", self.latlngs)
        print("WRITTEN GROUND TRUTH DATA TO CACHE")

    def write_var_to_cache(self, filename, var_name, data):
        f=open('Data/ground_truth/cache/'+var_name+'/'+filename+'.csv',"w+")
        for entry in data:
            line = str(entry[0, 0])+','+str(entry[0, 1])+','+str(entry[0, 2])+'\n'
#             line = str(entry[0])+','+str(entry[1])+','+str(entry[2])+'\n'

            f.write(line)
        f.close()