Mercurial Hosting > traffic-intelligence
view scripts/safety-analysis.py @ 946:e5970606066f
bug fix on list filtering (cannot remove while iterating) and motion prediction keeping the same features
author | Nicolas Saunier <nicolas.saunier@polymtl.ca> |
---|---|
date | Fri, 21 Jul 2017 11:25:20 -0400 |
parents | 05d4302bf67e |
children | 584b9405e494 |
line wrap: on
line source
#! /usr/bin/env python import storage, prediction, events, moving import sys, argparse, random import matplotlib.pyplot as plt import numpy as np # todo: very slow if too many predicted trajectories # add computation of probality of unsucessful evasive action parser = argparse.ArgumentParser(description='The program processes indicators for all pairs of road users in the scene') parser.add_argument('--cfg', dest = 'configFilename', help = 'name of the configuration file', required = True) parser.add_argument('-n', dest = 'nObjects', help = 'number of objects to analyse', type = int) # TODO analyze only parser.add_argument('--prediction-method', dest = 'predictionMethod', help = 'prediction method (constant velocity (cvd: vector computation (approximate); cve: equation solving; cv: discrete time (approximate)), normal adaptation, point set prediction)', choices = ['cvd', 'cve', 'cv', 'na', 'ps', 'mp']) parser.add_argument('--prototypeDatabaseFilename', dest = 'prototypeDatabaseFilename', help = 'name of the database containing the prototypes') parser.add_argument('--no-motion-prediction', dest = 'noMotionPrediction', help = 'does not compute indicators like TTC depending on motion prediction', action = 'store_true') parser.add_argument('--pet', dest = 'computePET', help = 'computes PET', action = 'store_true') parser.add_argument('--display-cp', dest = 'displayCollisionPoints', help = 'display collision points', action = 'store_true') parser.add_argument('--nthreads', dest = 'nProcesses', help = 'number of processes to run in parallel', type = int, default = 1) args = parser.parse_args() params = storage.ProcessParameters(args.configFilename) # parameters for prediction methods if args.predictionMethod is not None: predictionMethod = args.predictionMethod else: predictionMethod = params.predictionMethod def accelerationDistribution(): return random.triangular(-params.maxNormalAcceleration, params.maxNormalAcceleration, 0.) def steeringDistribution(): return random.triangular(-params.maxNormalSteering, params.maxNormalSteering, 0.) if predictionMethod == 'cvd': predictionParameters = prediction.CVDirectPredictionParameters() if predictionMethod == 'cve': predictionParameters = prediction.CVExactPredictionParameters() elif predictionMethod == 'cv': predictionParameters = prediction.ConstantPredictionParameters(params.maxPredictedSpeed) elif predictionMethod == 'na': predictionParameters = prediction.NormalAdaptationPredictionParameters(params.maxPredictedSpeed, params.nPredictedTrajectories, accelerationDistribution, steeringDistribution, params.useFeaturesForPrediction) elif predictionMethod == 'ps': predictionParameters = prediction.PointSetPredictionParameters(params.maxPredictedSpeed) elif predictionMethod == 'mp': if args.prototypeDatabaseFilename is None: prototypes = storage.loadPrototypesFromSqlite(params.databaseFilename) else: prototypes = storage.loadPrototypesFromSqlite(args.prototypeDatabaseFilename) for p in prototypes: p.getMovingObject().computeCumulativeDistances() predictionParameters = prediction.PrototypePredictionParameters(prototypes, params.nPredictedTrajectories, 2., 0.4, 'cityblock', 10, params.constantSpeedPrototypePrediction, params.useFeaturesForPrediction) # else: # no else required, since parameters is required as argument # evasiveActionPredictionParameters = prediction.EvasiveActionPredictionParameters(params.maxPredictedSpeed, # params.nPredictedTrajectories, # params.minExtremeAcceleration, # params.maxExtremeAcceleration, # params.maxExtremeSteering, # params.useFeaturesForPrediction) objects = storage.loadTrajectoriesFromSqlite(params.databaseFilename, 'object', args.nObjects, withFeatures = (params.useFeaturesForPrediction or predictionMethod == 'ps' or predictionMethod == 'mp')) # if params.useFeaturesForPrediction: # features = storage.loadTrajectoriesFromSqlite(params.databaseFilename,'feature') # needed if normal adaptation # for obj in objects: # obj.setFeatures(features) interactions = events.createInteractions(objects) for inter in interactions: inter.computeIndicators() if not args.noMotionPrediction: inter.computeCrossingsCollisions(predictionParameters, params.collisionDistance, params.predictionTimeHorizon, params.crossingZones, nProcesses = args.nProcesses, debug = True) if args.computePET: for inter in interactions: inter.computePET(params.collisionDistance) storage.saveIndicatorsToSqlite(params.databaseFilename, interactions) if args.displayCollisionPoints: plt.figure() allCollisionPoints = [] for inter in interactions: for collisionPoints in inter.collisionPoints.values(): allCollisionPoints += collisionPoints moving.Point.plotAll(allCollisionPoints) plt.axis('equal')