Mercurial Hosting > traffic-intelligence
comparison scripts/classify-objects.py @ 902:c69a8defe5c3
changed workflow of classify objects
author | Nicolas Saunier <nicolas.saunier@polymtl.ca> |
---|---|
date | Thu, 22 Jun 2017 16:57:34 -0400 |
parents | 753a081989e2 |
children | 8f60ecfc2f06 |
comparison
equal
deleted
inserted
replaced
901:753a081989e2 | 902:c69a8defe5c3 |
---|---|
18 parser.add_argument('--plot-speed-distributions', dest = 'plotSpeedDistribution', help = 'simply plots the distributions used for each user type', action = 'store_true') | 18 parser.add_argument('--plot-speed-distributions', dest = 'plotSpeedDistribution', help = 'simply plots the distributions used for each user type', action = 'store_true') |
19 parser.add_argument('--max-speed-distribution-plot', dest = 'maxSpeedDistributionPlot', help = 'if plotting the user distributions, the maximum speed to display (km/h)', type = float, default = 50.) | 19 parser.add_argument('--max-speed-distribution-plot', dest = 'maxSpeedDistributionPlot', help = 'if plotting the user distributions, the maximum speed to display (km/h)', type = float, default = 50.) |
20 | 20 |
21 args = parser.parse_args() | 21 args = parser.parse_args() |
22 params, videoFilename, databaseFilename, invHomography, intrinsicCameraMatrix, distortionCoefficients, undistortedImageMultiplication, undistort, firstFrameNum = storage.processVideoArguments(args) | 22 params, videoFilename, databaseFilename, invHomography, intrinsicCameraMatrix, distortionCoefficients, undistortedImageMultiplication, undistort, firstFrameNum = storage.processVideoArguments(args) |
23 | |
24 classifierParams = storage.ClassifierParameters(params.classifierFilename) | 23 classifierParams = storage.ClassifierParameters(params.classifierFilename) |
25 classifierParams.convertToFrames(params.videoFrameRate, 3.6) # conversion from km/h to m/frame | 24 classifierParams.convertToFrames(params.videoFrameRate, 3.6) # conversion from km/h to m/frame |
26 | 25 |
27 if classifierParams.speedAggregationMethod == 'median': | 26 if classifierParams.speedAggregationMethod == 'median': |
28 speedAggregationFunc = np.median | 27 speedAggregationFunc = np.median |
64 plt.title('Probability Density Function') | 63 plt.title('Probability Density Function') |
65 plt.show() | 64 plt.show() |
66 sys.exit() | 65 sys.exit() |
67 | 66 |
68 objects = storage.loadTrajectoriesFromSqlite(databaseFilename, 'object', args.nObjects, withFeatures = True) | 67 objects = storage.loadTrajectoriesFromSqlite(databaseFilename, 'object', args.nObjects, withFeatures = True) |
69 #features = storage.loadTrajectoriesFromSqlite(databaseFilename, 'feature') | 68 timeInterval = moving.TimeInterval.unionIntervals([obj.getTimeInterval() for obj in objects]) |
70 intervals = [] | |
71 for obj in objects: | |
72 #obj.setFeatures(features) | |
73 intervals.append(obj.getTimeInterval()) | |
74 timeInterval = moving.TimeInterval.unionIntervals(intervals) | |
75 | 69 |
76 capture = cv2.VideoCapture(videoFilename) | 70 capture = cv2.VideoCapture(videoFilename) |
77 width = int(capture.get(cv2.cv.CV_CAP_PROP_FRAME_WIDTH)) | 71 width = int(capture.get(cv2.cv.CV_CAP_PROP_FRAME_WIDTH)) |
78 height = int(capture.get(cv2.cv.CV_CAP_PROP_FRAME_HEIGHT)) | 72 height = int(capture.get(cv2.cv.CV_CAP_PROP_FRAME_HEIGHT)) |
79 | 73 |
80 pastObjects = [] | 74 pastObjects = [] |
75 currentObjects = [] | |
81 if undistort: # setup undistortion | 76 if undistort: # setup undistortion |
82 [map1, map2] = cvutils.computeUndistortMaps(width, height, undistortedImageMultiplication, intrinsicCameraMatrix, distortionCoefficients) | 77 [map1, map2] = cvutils.computeUndistortMaps(width, height, undistortedImageMultiplication, intrinsicCameraMatrix, distortionCoefficients) |
83 if capture.isOpened(): | 78 if capture.isOpened(): |
84 ret = True | 79 ret = True |
85 frameNum = timeInterval.first | 80 frameNum = timeInterval.first |
90 ret, img = capture.read() | 85 ret, img = capture.read() |
91 if ret: | 86 if ret: |
92 if frameNum%50 == 0: | 87 if frameNum%50 == 0: |
93 print('frame number: {}'.format(frameNum)) | 88 print('frame number: {}'.format(frameNum)) |
94 if undistort: | 89 if undistort: |
95 img = cv2.remap(img, map1, map2, interpolation=cv2.INTER_LINEAR) | 90 img = cv2.remap(img, map1, map2, interpolation=cv2.INTER_LINEAR) |
96 currentObjects = [] | |
97 for obj in objects: | 91 for obj in objects: |
98 inter = obj.getTimeInterval() | 92 if obj.getFirstInstant() == frameNum: |
99 if inter.contains(frameNum): | 93 obj.initClassifyUserTypeHoGSVM(speedAggregationFunc, pedBikeCarSVM, bikeCarSVM, classifierParams.maxPedestrianSpeed, classifierParams.maxCyclistSpeed, classifierParams.nFramesIgnoreAtEnds) |
100 if inter.first == frameNum: | 94 currentObjects.append(obj) |
101 obj.initClassifyUserTypeHoGSVM(speedAggregationFunc, pedBikeCarSVM, bikeCarSVM, classifierParams.maxPedestrianSpeed, classifierParams.maxCyclistSpeed, classifierParams.nFramesIgnoreAtEnds) | 95 objects.remove(obj) |
102 currentObjects.append(obj) | 96 |
103 elif inter.last == frameNum: | 97 for obj in currentObjects: |
104 obj.classifyUserTypeHoGSVM(minSpeedEquiprobable = classifierParams.minSpeedEquiprobable, speedProbabilities = speedProbabilities, maxPercentUnknown = classifierParams.maxPercentUnknown) | 98 if obj.getLastInstant() == frameNum: |
105 pastObjects.append(obj) | 99 obj.classifyUserTypeHoGSVM(minSpeedEquiprobable = classifierParams.minSpeedEquiprobable, speedProbabilities = speedProbabilities, maxPercentUnknown = classifierParams.maxPercentUnknown) |
106 else: | 100 pastObjects.append(obj) |
107 obj.classifyUserTypeHoGSVMAtInstant(img, frameNum, invHomography, width, height, classifierParams.percentIncreaseCrop, classifierParams.percentIncreaseCrop, classifierParams.minNPixels, classifierParams.hogRescaleSize, classifierParams.hogNOrientations, classifierParams.hogNPixelsPerCell, classifierParams.hogNCellsPerBlock, classifierParams.hogBlockNorm) | 101 currentObjects.remove(obj) |
108 currentObjects.append(obj) | |
109 else: | 102 else: |
110 currentObjects.append(obj) | 103 obj.classifyUserTypeHoGSVMAtInstant(img, frameNum, invHomography, width, height, classifierParams.percentIncreaseCrop, classifierParams.percentIncreaseCrop, classifierParams.minNPixels, classifierParams.hogRescaleSize, classifierParams.hogNOrientations, classifierParams.hogNPixelsPerCell, classifierParams.hogNCellsPerBlock, classifierParams.hogBlockNorm) |
111 objects = currentObjects | |
112 frameNum += 1 | 104 frameNum += 1 |
113 | 105 |
114 for obj in objects: | 106 for obj in currentObjects: |
115 obj.classifyUserTypeHoGSVM(minSpeedEquiprobable = classifierParams.minSpeedEquiprobable, speedProbabilities = speedProbabilities, maxPercentUnknown = classifierParams.maxPercentUnknown) | 107 obj.classifyUserTypeHoGSVM(minSpeedEquiprobable = classifierParams.minSpeedEquiprobable, speedProbabilities = speedProbabilities, maxPercentUnknown = classifierParams.maxPercentUnknown) |
116 pastObjects.append(obj) | 108 pastObjects.append(obj) |
117 print('Saving user types') | 109 print('Saving user types') |
118 storage.setRoadUserTypes(databaseFilename, pastObjects) | 110 storage.setRoadUserTypes(databaseFilename, pastObjects) |