Mercurial Hosting > traffic-intelligence

comparison scripts/classify-objects.py @ 708:a37c565f4b68

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
merged dev
author Nicolas Saunier <nicolas.saunier@polymtl.ca>
date Wed, 22 Jul 2015 14:17:44 -0400
parents de278c5e65f6
children 5b970a5bc233
comparison
equal deleted inserted replaced
707:7efa36b9bcfd 708:a37c565f4b68
1 #! /usr/bin/env python 1 #! /usr/bin/env python
2
3 import cvutils, moving, ml, storage
2 4
3 import numpy as np 5 import numpy as np
4 import sys, argparse 6 import sys, argparse
5 from cv2 import SVM_RBF, SVM_C_SVC 7 #from cv2 import SVM_RBF, SVM_C_SVC
8 import cv2
9 from scipy.stats import norm, lognorm
6 10
7 import cvutils, moving, ml 11 # TODO add mode detection live
8 12
9 parser = argparse.ArgumentParser(description='The program processes indicators for all pairs of road users in the scene') 13 parser = argparse.ArgumentParser(description='The program processes indicators for all pairs of road users in the scene')
10 parser.add_argument('--cfg', dest = 'configFilename', help = 'name of the configuration file', required = True) 14 parser.add_argument('--cfg', dest = 'configFilename', help = 'name of the configuration file', required = True)
11 parser.add_argument('-d', dest = 'directoryName', help = 'name of the parent directory containing the videos and extracted trajectories to process', required = True) 15 parser.add_argument('-d', dest = 'databaseFilename', help = 'name of the Sqlite database file (overrides the configuration file)')
12 # parser.add_argument('-o', dest = 'homographyFilename', help = 'name of the image to world homography file') 16 parser.add_argument('-i', dest = 'videoFilename', help = 'name of the video file (overrides the configuration file)')
13 # need a classification config file for speed distribution parameters, svm models, frequency parameters, area parameters etc 17 parser.add_argument('-n', dest = 'nObjects', help = 'number of objects to classify', type = int, default = None)
14 #parser.add_argument('--cfg', dest = 'svmType', help = 'SVM type', default = SVM_C_SVC, type = long) 18 parser.add_argument('--plot-speed-distributions', dest = 'plotSpeedDistribution', help = 'simply plots the distributions used for each user type', action = 'store_true')
15 19 parser.add_argument('--max-speed-distribution-plot', dest = 'maxSpeedDistributionPlot', help = 'if plotting the user distributions, the maximum speed to display', type = float, default = 50.)
16
17 #parser.add_argument('-s', dest = 'rescaleSize', help = 'rescale size of image samples', default = 64, type = int)
18 #parser.add_argument('-o', dest = 'nOrientations', help = 'number of orientations in HoG', default = 9, type = int)
19 #parser.add_argument('-p', dest = 'nPixelsPerCell', help = 'number of pixels per cell', default = 8, type = int)
20 #parser.add_argument('-c', dest = 'nCellsPerBlock', help = 'number of cells per block', default = 2, type = int)
21 20
22 args = parser.parse_args() 21 args = parser.parse_args()
23 params = storage.ProcessParameters(args.configFilename) 22 params = storage.ProcessParameters(args.configFilename)
23
24 if args.videoFilename is not None:
25 videoFilename = args.videoFilename
26 else:
27 videoFilename = params.videoFilename
28 if args.databaseFilename is not None:
29 databaseFilename = args.databaseFilename
30 else:
31 databaseFilename = params.databaseFilename
32
33 params.convertToFrames(3.6)
34 if params.homography is not None:
35 invHomography = np.linalg.inv(params.homography)
36
37 if params.speedAggregationMethod == 'median':
38 speedAggregationFunc = np.median
39 elif params.speedAggregationMethod == 'mean':
40 speedAggregationFunc = np.mean
41 elif params.speedAggregationMethod == 'quantile':
42 speedAggregationFunc = lambda speeds: np.percentile(speeds, args.speedAggregationQuantile)
43 else:
44 print('Unknown speed aggregation method: {}. Exiting'.format(params.speedAggregationMethod))
45 sys.exit()
46
47 pedBikeCarSVM = ml.SVM()
48 pedBikeCarSVM.load(params.pedBikeCarSVMFilename)
49 bikeCarSVM = ml.SVM()
50 bikeCarSVM.load(params.bikeCarSVMFilename)
51
52 # log logistic for ped and bik otherwise ((pedBeta/pedAlfa)*((sMean/pedAlfa)**(pedBeta-1)))/((1+(sMean/pedAlfa)**pedBeta)**2.)
53 speedProbabilities = {'car': lambda s: norm(params.meanVehicleSpeed, params.stdVehicleSpeed).pdf(s),
54 'pedestrian': lambda s: norm(params.meanPedestrianSpeed, params.stdPedestrianSpeed).pdf(s),
55 'bicycle': lambda s: lognorm(params.scaleCyclistSpeed, loc = 0., scale = np.exp(params.locationCyclistSpeed)).pdf(s)} # numpy lognorm shape, loc, scale: shape for numpy is scale (std of the normal) and scale for numpy is location (mean of the normal)
56
57 if args.plotSpeedDistribution:
58 import matplotlib.pyplot as plt
59 plt.figure()
60 for k in speedProbabilities:
61 plt.plot(np.arange(0.1, args.maxSpeedDistributionPlot, 0.1), [speedProbabilities[k](s/3.6/25) for s in np.arange(0.1, args.maxSpeedDistributionPlot, 0.1)], label = k)
62 plt.xlabel('Speed (km/h)')
63 plt.ylabel('Probability')
64 plt.legend()
65 plt.title('Probability Density Function')
66 plt.show()
67 sys.exit()
68
69 objects = storage.loadTrajectoriesFromSqlite(databaseFilename, 'object', args.nObjects, withFeatures = True)
70 #features = storage.loadTrajectoriesFromSqlite(databaseFilename, 'feature')
71 intervals = []
72 for obj in objects:
73 #obj.setFeatures(features)
74 intervals.append(obj.getTimeInterval())
75 timeInterval = moving.unionIntervals(intervals)
76
77 capture = cv2.VideoCapture(videoFilename)
78 width = int(capture.get(cv2.cv.CV_CAP_PROP_FRAME_WIDTH))
79 height = int(capture.get(cv2.cv.CV_CAP_PROP_FRAME_HEIGHT))
80
81 pastObjects = []
82 if params.undistort: # setup undistortion
83 [map1, map2] = computeUndistortMaps(width, height, undistortedImageMultiplication, intrinsicCameraMatrix, distortionCoefficients)
84 if capture.isOpened():
85 ret = True
86 frameNum = timeInterval.first
87 capture.set(cv2.cv.CV_CAP_PROP_POS_FRAMES, frameNum)
88 lastFrameNum = timeInterval.last
89
90 while ret and frameNum <= lastFrameNum:
91 ret, img = capture.read()
92 if ret:
93 if frameNum%50 == 0:
94 print('frame number: {}'.format(frameNum))
95 currentObjects = []
96 for obj in objects:
97 if obj.getLastInstant() < frameNum:
98 obj.classifyUserTypeHoGSVM(minSpeedEquiprobable = params.minSpeedEquiprobable, speedProbabilities = speedProbabilities)
99 pastObjects.append(obj)
100 else:
101 currentObjects.append(obj)
102 objects = currentObjects
103 if params.undistort:
104 img = cv2.remap(img, map1, map2, interpolation=cv2.INTER_LINEAR)
105 for obj in objects:
106 if obj.existsAtInstant(frameNum):
107 if obj.getFirstInstant() == frameNum:
108 obj.initClassifyUserTypeHoGSVM(speedAggregationFunc, pedBikeCarSVM, bikeCarSVM, params.maxPedestrianSpeed, params.maxCyclistSpeed, params.nFramesIgnoreAtEnds)
109 obj.classifyUserTypeHoGSVMAtInstant(img, frameNum, invHomography, width, height, 0.2, 0.2, 800) # px, py, pixelThreshold
110 frameNum += 1
111
112 for obj in objects:
113 obj.classifyUserTypeHoGSVM(minSpeedEquiprobable = params.minSpeedEquiprobable, speedProbabilities = speedProbabilities)
114 pastObjects.append(obj)
115 print('Saving user types')
116 storage.setRoadUserTypes(databaseFilename, pastObjects)