Mercurial > hg > nsaunier > traffic-intelligence

comparison scripts/process.py @ 1046:f2ba9858e6c6

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
motion pattern learning seems to work
author Nicolas Saunier <nicolas.saunier@polymtl.ca>
date Thu, 05 Jul 2018 23:12:11 -0400
parents 25db2383e7ae
children 0b62e37991ab
comparison
equal deleted inserted replaced
1045:25db2383e7ae 1046:f2ba9858e6c6
5 from multiprocessing.pool import Pool 5 from multiprocessing.pool import Pool
6 6
7 #import matplotlib 7 #import matplotlib
8 #atplotlib.use('Agg') 8 #atplotlib.use('Agg')
9 import matplotlib.pyplot as plt 9 import matplotlib.pyplot as plt
10 from numpy import percentile 10 from numpy import percentile, ones
11 from pandas import DataFrame 11 from pandas import DataFrame
12 12
13 from trafficintelligence import storage, events, prediction, cvutils, utils, moving 13 from trafficintelligence import storage, events, prediction, cvutils, utils, moving, processing, ml
14 from trafficintelligence.metadata import * 14 from trafficintelligence.metadata import *
15 15
16 parser = argparse.ArgumentParser(description='This program manages the processing of several files based on a description of the sites and video data in an SQLite database following the metadata module.') 16 parser = argparse.ArgumentParser(description='This program manages the processing of several files based on a description of the sites and video data in an SQLite database following the metadata module.')
17 # input 17 # input
18 parser.add_argument('--db', dest = 'metadataFilename', help = 'name of the metadata file', required = True) 18 parser.add_argument('--db', dest = 'metadataFilename', help = 'name of the metadata file', required = True)
19 parser.add_argument('--videos', dest = 'videoIds', help = 'indices of the video sequences', nargs = '*', type = int) 19 parser.add_argument('--videos', dest = 'videoIds', help = 'indices of the video sequences', nargs = '*', type = int)
20 parser.add_argument('--sites', dest = 'siteIds', help = 'indices of the video sequences', nargs = '*', type = int) 20 parser.add_argument('--sites', dest = 'siteIds', help = 'indices of the video sequences', nargs = '*')
21 21
22 # main function 22 # main function
23 parser.add_argument('--delete', dest = 'delete', help = 'data to delete', choices = ['feature', 'object', 'classification', 'interaction']) 23 parser.add_argument('--delete', dest = 'delete', help = 'data to delete', choices = ['feature', 'object', 'classification', 'interaction'])
24 parser.add_argument('--process', dest = 'process', help = 'data to process', choices = ['feature', 'object', 'classification', 'prototype', 'interaction']) 24 parser.add_argument('--process', dest = 'process', help = 'data to process', choices = ['feature', 'object', 'classification', 'prototype', 'interaction'])
25 parser.add_argument('--display', dest = 'display', help = 'data to display (replay over video)', choices = ['feature', 'object', 'classification', 'interaction']) 25 parser.add_argument('--display', dest = 'display', help = 'data to display (replay over video)', choices = ['feature', 'object', 'classification', 'interaction'])
30 parser.add_argument('-n', dest = 'nObjects', help = 'number of objects/interactions to process', type = int) 30 parser.add_argument('-n', dest = 'nObjects', help = 'number of objects/interactions to process', type = int)
31 parser.add_argument('-t', dest = 'trajectoryType', help = 'type of trajectories', choices = ['feature', 'object'], default = 'feature') 31 parser.add_argument('-t', dest = 'trajectoryType', help = 'type of trajectories', choices = ['feature', 'object'], default = 'feature')
32 parser.add_argument('--dry', dest = 'dryRun', help = 'dry run of processing', action = 'store_true') 32 parser.add_argument('--dry', dest = 'dryRun', help = 'dry run of processing', action = 'store_true')
33 parser.add_argument('--nthreads', dest = 'nProcesses', help = 'number of processes to run in parallel', type = int, default = 1) 33 parser.add_argument('--nthreads', dest = 'nProcesses', help = 'number of processes to run in parallel', type = int, default = 1)
34 parser.add_argument('--subsample', dest = 'positionSubsamplingRate', help = 'rate of position subsampling (1 every n positions)', type = int) 34 parser.add_argument('--subsample', dest = 'positionSubsamplingRate', help = 'rate of position subsampling (1 every n positions)', type = int)
35 parser.add_argument('--display', dest = 'display', help = 'display trajectories', action = 'store_true')
36 35
37 ### process options 36 ### process options
38 # motion pattern learning and assignment 37 # motion pattern learning and assignment
39 parser.add_argument('--prototype-filename', dest = 'outputPrototypeDatabaseFilename', help = 'name of the Sqlite database file to save prototypes') 38 parser.add_argument('--prototype-filename', dest = 'outputPrototypeDatabaseFilename', help = 'name of the Sqlite database file to save prototypes', default = 'prototypes.sqlite')
40 #parser.add_argument('-i', dest = 'inputPrototypeDatabaseFilename', help = 'name of the Sqlite database file for prototypes to start the algorithm with') 39 #parser.add_argument('-i', dest = 'inputPrototypeDatabaseFilename', help = 'name of the Sqlite database file for prototypes to start the algorithm with')
41 parser.add_argument('--nfeatures-per-object', dest = 'nLongestFeaturesPerObject', help = 'maximum number of features per object to load', type = int) 40 parser.add_argument('--nfeatures-per-object', dest = 'nLongestFeaturesPerObject', help = 'maximum number of features per object to load', type = int)
42 parser.add_argument('--epsilon', dest = 'epsilon', help = 'distance for the similarity of trajectory points', type = float, required = True) 41 parser.add_argument('--epsilon', dest = 'epsilon', help = 'distance for the similarity of trajectory points', type = float)
43 parser.add_argument('--metric', dest = 'metric', help = 'metric for the similarity of trajectory points', default = 'cityblock') # default is manhattan distance 42 parser.add_argument('--metric', dest = 'metric', help = 'metric for the similarity of trajectory points', default = 'cityblock') # default is manhattan distance
44 parser.add_argument('--minsimil', dest = 'minSimilarity', help = 'minimum similarity to put a trajectory in a cluster', type = float, required = True) 43 parser.add_argument('--minsimil', dest = 'minSimilarity', help = 'minimum similarity to put a trajectory in a cluster', type = float)
45 parser.add_argument('-min-cluster-size', dest = 'minClusterSize', help = 'minimum cluster size', type = int, default = 0) 44 parser.add_argument('--min-cluster-size', dest = 'minClusterSize', help = 'minimum cluster size', type = int, default = 0)
46 #parser.add_argument('--learn', dest = 'learn', help = 'learn', action = 'store_true') 45 #parser.add_argument('--learn', dest = 'learn', help = 'learn', action = 'store_true')
47 parser.add_argument('--optimize', dest = 'optimizeCentroid', help = 'recompute centroid at each assignment', action = 'store_true') 46 parser.add_argument('--optimize', dest = 'optimizeCentroid', help = 'recompute centroid at each assignment', action = 'store_true')
48 parser.add_argument('--random', dest = 'randomInitialization', help = 'random initialization of clustering algorithm', action = 'store_true') 47 parser.add_argument('--random', dest = 'randomInitialization', help = 'random initialization of clustering algorithm', action = 'store_true')
49 #parser.add_argument('--similarities-filename', dest = 'similaritiesFilename', help = 'filename of the similarities') 48 #parser.add_argument('--similarities-filename', dest = 'similaritiesFilename', help = 'filename of the similarities')
50 parser.add_argument('--save-similarities', dest = 'saveSimilarities', help = 'save computed similarities (in addition to prototypes)', action = 'store_true') 49 parser.add_argument('--save-similarities', dest = 'saveSimilarities', help = 'save computed similarities (in addition to prototypes)', action = 'store_true')
153 objects = {} 152 objects = {}
154 object2VideoSequences = {} 153 object2VideoSequences = {}
155 for cv in site.cameraViews: 154 for cv in site.cameraViews:
156 for vs in cv.videoSequences: 155 for vs in cv.videoSequences:
157 print('Loading '+vs.getDatabaseFilename()) 156 print('Loading '+vs.getDatabaseFilename())
158 objects[vs.idx] = storage.loadTrajectoriesFromSqlite(str(parentPath/vs.getDatabaseFilename()), args.trajectoryType, args.nTrajectories, timeStep = args.positionSubsamplingRate, nLongestFeaturesPerObject = args.nLongestFeaturesPerObject) 157 objects[vs.idx] = storage.loadTrajectoriesFromSqlite(str(parentPath/vs.getDatabaseFilename()), args.trajectoryType, args.nObjects, timeStep = args.positionSubsamplingRate, nLongestFeaturesPerObject = args.nLongestFeaturesPerObject)
159 if args.trajectoryType == 'object' and args.nLongestFeaturesPerObject is not None: 158 if args.trajectoryType == 'object' and args.nLongestFeaturesPerObject is not None:
160 objectsWithFeatures = objects[vs.idx] 159 objectsWithFeatures = objects[vs.idx]
161 objects[vs.idx] = [f for o in objectsWithFeatures for f in o.getFeatures()] 160 objects[vs.idx] = [f for o in objectsWithFeatures for f in o.getFeatures()]
162 prototypeType = 'feature' 161 prototypeType = 'feature'
163 else: 162 else:
165 for obj in objects[vs.idx]: 164 for obj in objects[vs.idx]:
166 object2VideoSequences[obj] = vs 165 object2VideoSequences[obj] = vs
167 lcss = utils.LCSS(metric = args.metric, epsilon = args.epsilon) 166 lcss = utils.LCSS(metric = args.metric, epsilon = args.epsilon)
168 similarityFunc = lambda x,y : lcss.computeNormalized(x, y) 167 similarityFunc = lambda x,y : lcss.computeNormalized(x, y)
169 allobjects = [o for tmpobjects in objects.values() for o in tmpobjects] 168 allobjects = [o for tmpobjects in objects.values() for o in tmpobjects]
170 prototypeIndices, labels = processing.learnAssignMotionPatterns(True, True, allobjects, similarities, args.minsimil, similarityFunc, args.minClusterSize, args.optimizeCentroid, args.randomInitialization, True, []) 169 similarities = -ones((len(allobjects), len(allobjects)))
170 prototypeIndices, labels = processing.learnAssignMotionPatterns(True, True, allobjects, similarities, args.minSimilarity, similarityFunc, args.minClusterSize, args.optimizeCentroid, args.randomInitialization, True, [])
171 if args.outputPrototypeDatabaseFilename is None: 171 if args.outputPrototypeDatabaseFilename is None:
172 outputPrototypeDatabaseFilename = args.databaseFilename 172 outputPrototypeDatabaseFilename = args.databaseFilename
173 else: 173 else:
174 outputPrototypeDatabaseFilename = args.outputPrototypeDatabaseFilename 174 outputPrototypeDatabaseFilename = args.outputPrototypeDatabaseFilename
175 # TODO maintain mapping from object prototype to db filename + compute nmatchings before 175 # TODO maintain mapping from object prototype to db filename + compute nmatchings before
176 clusterSizes = ml.computeClusterSizes(labels, prototypeIndices, -1) 176 clusterSizes = ml.computeClusterSizes(labels, prototypeIndices, -1)
177 storage.savePrototypesToSqlite(outputPrototypeDatabaseFilename, [moving.Prototype(object2VideoSequences[allobjects[i]].getDatabaseFilename(False), allobjects[i].getNum(), prototypeType) for i in prototypeIndices]) 177 storage.savePrototypesToSqlite(str(parentPath/site.getPath()/outputPrototypeDatabaseFilename), [moving.Prototype(object2VideoSequences[allobjects[i]].getDatabaseFilename(False), allobjects[i].getNum(), prototypeType, clusterSizes[i]) for i in prototypeIndices])
178 178
179 179
180 elif args.process == 'interaction': 180 elif args.process == 'interaction':
181 # safety analysis TODO make function in safety analysis script 181 # safety analysis TODO make function in safety analysis script
182 if args.predictionMethod == 'cvd': 182 if args.predictionMethod == 'cvd':