A methodology for monitoring traffjc A methodology for monitoring - - PowerPoint PPT Presentation

▶
a methodology for monitoring traffjc a methodology for
SMART_READER_LITE
LIVE PREVIEW

A methodology for monitoring traffjc A methodology for monitoring - - PowerPoint PPT Presentation

Jul 20, 2023 238 likes •525 views

A methodology for monitoring traffjc A methodology for monitoring traffjc fmow and air pollution in urban areas fmow and air pollution in urban areas Jos ngel Martn-Baos Jos ngel Martn-Baos SYSORM 2019 SYSORM 2019 Ricardo

slide-1
SLIDE 1

A methodology for monitoring traffjc A methodology for monitoring traffjc fmow and air pollution in urban areas fmow and air pollution in urban areas

José Ángel Martín-Baos José Ángel Martín-Baos Ricardo García-Ródenas Ricardo García-Ródenas Luis Rodríguez-Benitez Luis Rodríguez-Benitez SYSORM 2019 SYSORM 2019 5 5th

th June 2019

June 2019

slide-2
SLIDE 2

2

Content Content

Introductjon Conclusions Results Methodology Research Questjon

05 04 03 02 01

slide-3
SLIDE 3

Operational level Operational level

Real traffic control Real traffic control

3

Tactic level Tactic level

Capacity enhancement problem Capacity enhancement problem

Strategic level Strategic level

Traffic Network Layout Traffic Network Layout Signal setting problem Signal setting problem Toll pricing problem Toll pricing problem Demand adjustment problem Demand adjustment problem

Introduction Introduction

slide-4
SLIDE 4

Operational level Operational level

Real traffic control Real traffic control

4

Technologies Technologies Methodology Methodology

Proyect awarded by:

slide-5
SLIDE 5
  • 5

Research Question Research Question

Emission monitoring:

  • Low cost devices
  • Distributed environment
slide-6
SLIDE 6

Source: European Environment Agency

Years % change relative to 1990

Traffic pollution Traffic pollution in Europe in Europe

6

slide-7
SLIDE 7

Source: www.cocheando.es

Environmental Environmental labels in Spain labels in Spain

7

slide-8
SLIDE 8

Methodology Methodology

8

slide-9
SLIDE 9

9 Source: www.raspberrypi.org

Raspberry Pi Raspberry Pi

slide-10
SLIDE 10

10

Motion Vectors Motion Vectors

slide-11
SLIDE 11

11

Obtain the motion Obtain the motion vectors vectors

slide-12
SLIDE 12

12

Obtain the motion Obtain the motion vectors vectors

slide-13
SLIDE 13

13

Simple moving Simple moving average average

slide-14
SLIDE 14

14

Algorithm to Algorithm to count vehicles count vehicles

n_vehicles ← 1 car_detected False ← 1 growth ← 1 GROWTH_LIMIT 5 ← 1 WIDTH_THRESHOLD 10 ← 1 HEIGHT_THRESHOLD 150 ← 1 SMOOTH_ORDER 6 ← 1 /* Repeat for each frame in the video */ mv ← 1 smooth(mv, SMOOTH_ORDER)

slide-15
SLIDE 15

15

Algorithm to Algorithm to count vehicles count vehicles

/* Repeat for Left and Rigth direction */ if mv[previous frame] < mv and growth < GROWTH_LIMIT then growth growth ← 1 + 1; else if mv[previous frame] > mv and growth > - GROWTH_LIMIT then growth growth ← 1

  • 1;

end if mv >= HEIGHT_THRESHOLD and growth > 0 then n_positive_frames n_positive_frames ← 1 + 1; if n_positive_frames >= WIDTH_THRESHOLD and car_detected = False then car_detected True; ← 1 n_vehicles n_vehicles ← 1 + 1; end else if growth = - GROWTH_LIMIT then car_detected False; ← 1 n_positive_frames ← 1 0; end

slide-16
SLIDE 16

16

DEMO

Algorithm to Algorithm to count vehicles count vehicles

slide-17
SLIDE 17

17

Connection Connection diagram diagram

slide-18
SLIDE 18

18

Raspberry Pi Sensor Server Raspberry Pi Sensor

MQTT Broker

Publish: "24 ºC" T

  • pic: "T

emperature" Subscribe to: "T emperature" Publish: "24 ºC", Publish: "22.5 ºC" Publish: "22.5 ºC" T

  • pic: "T

emperature"

IBM IoT IBM IoT Platform Platform

slide-19
SLIDE 19

19

Results Results

slide-20
SLIDE 20

20

~ 9 ~ 90 % % average percentage hits

using 23 test videos in a street with

  • ne line in each direction

0,4% of the information used

compared to traditional video analysis techniques

~ 0 ~ 0,0 ,0024 ~ 0 ~ 0,0 ,0024 seconds per frame

slide-21
SLIDE 21

21

Web Page Web Page

slide-22
SLIDE 22

22

Web Page Web Page

slide-23
SLIDE 23

23

Web Page Web Page

slide-24
SLIDE 24

24

Web Page Web Page

slide-25
SLIDE 25

25

Conclusion Conclusion

Motion vectors allows to use statistical techniques to detect vehicles without processing the image itself. ~90% of vehicles detected. Low CPU resources (~10%). Fast (~0,0024 s/frame)

ALGORITHM

We are able to monitor simultaneously road traffic and air pollution using a low-cost device.

MONITOR

Implement a machine learning based methodology to estimate the air quality evolution using the data provided by this infrastructure and recommend palliative actions.

FUTURE WORK

We are able to communicate all the information to the cloud using IBM services, were it can be processed.

COMUNICATION

slide-26
SLIDE 26
slide-27
SLIDE 27

José Ángel Martîn Baos

Universidad de Castjlla-La Mancha

JoseAngel.Martjn@uclm.es

Thank you for your attention

htups://linkedin.com/in/joseangelmartjnb/