Moving Objects and Spatial Data Computing

Moving Objects and
Spatial Data Computing
(이동체 및 공간 데이터 컴퓨팅 연구 동향)
2017. 09. 28
Kwang Woo NAM
kwnam@kunsan.ac.kr
Kunsan National University
This research was supported by a Grant (14NSIP‐B080144‐01) from National Land Space Information
Research Program funded by Ministry of Land, Infrastructure and Transport of Korean government. 2017.09.28 KAIST 조천식녹색교통대학원 발표자료

2
Who
남광우 교수/ 국립 군산대학교
‐ 군산대 컴퓨터정보통신공학부(2004‐현재)
‐ 미래형자동차 R&D 전문인력양성사업단(2017‐)
{ KEA, 군산대, 한양대, 충북대, 인하대, 부품연}
‐ 前 ETRI 텔레매틱스연구단 선임연구원(2001‐2004)
‐ U. of Minnesota Visiting Scholar(2015‐2016)
Projects
• 국토해양부, 공간정보 S/W활용을 위한 오픈소스 가공기술 개발(14‐19) : PostTrajectory
• NRF, 소셜 미디어 스트림 기반 공간 지식의 연속 마이닝(13‐16): SKSpark, SpatialTajo
• ETRI, 시멘틱 공간정보 서비스 프레임워크 및 구조 분석(10)
• 국토해양부, U‐GIS 국토정보 컨텐츠 서비스 요소 기술 연구(07‐11)
• NRF, geoSensor 센서 스트림 기반 점진적 온라인 패턴 마이닝 기법 연구(09‐12)
• NRF, GeoSensor Network를 위한 효율적인 공간 센서 스트림 죠인 및 집계처리 기법 연구(06‐07)
• NRF, 상황정보 기반의 Networked GeoVision 기술 개발(08‐11) : GeoVIsion

3
• Spatial Data
• 지상, 지하, 수상, 공중 등의 공간 상에 존재하는 자연적 또
는 인공적 객체들의 위치정보와 속성정보 및 객체 들간의
위상정보
Introduction
(a) GIS (b) BIM(CAD)

4
• Spatial Information vs. Geographic Information
Introduction
Spatial Information
Geographic
(GeoSpatial)
Information
Location
Information
Space
Information
CAD
Information
Indoor
Information

Introduction
공간 데이터
뉴미디어 공간정보 콘텐츠
공간 정보
공간정보 Spectrum의 확대
2D Vector
RS Data
2.5D Data DEM Data
LBS/TelematicsGeoERPWeb GIS
MultiSensor
Smart Phone
Real 3D Data 360 Panorama
Video+GPS
(blackbox)
Game,Edu,…
Closed Box(Public Sector)
Open Box( Enterpise Sector)
Shared Box(People Sector)
Sony Camera
(GPS+Pano) iWatch
Google Glass
GoPro
• Evolutions of Spatial Data/Information

Introduction
• 일반적인 공간좌표 및 POI 등의 공간 정보와 웹 문서, 도서, 사진,
동영상, 음성 등의 미디어가 결합된 공간정보 컨텐츠들의 등장
• 예) GoPro, PointCloud, 차량 블랙박스, 360 파노라마, 구글 글래스
GeoTagged Photo(Panoramio) 360 Panorama(StreetView)
Car Blackbox Video 자율주행차 PointCloud(LiDAR)

7
• Converging into Spatial Data
Introduction
GNSS - 무선데이터통신망을
이용한 정밀위치정보제공
- 차량항법, 교통정보 제공
GIS SIIS
ITS
- 정밀 기준점 위치정보에
의한 영상정보 정확도 향상
- GPS 기지국 구축적지 선정
- 차량용 네비게이션 정보제공
- 교통주제도 구축 및 계량적
분석 기술제공
- 교통시설물 관리 및 IT기반
운영기술 제공
- 대용량 공간정보
자동수정 갱신
- 영상지도 제작
- 측량기술 및 관련기술
(위성삼각측량, WGS84)제공
- 실시간 Mobile GIS응용기술
- 영상기반 교통정보
수집기술제공
- 고 정밀 영상검지
시스템 구축
출처 : ETRI

공간정보 서비스 개발 플랫폼

9
• 버스정보시스템과 스마트폰 앱
Spatial Data Services
버스정보시스템 주변 맛집 앱
정재준, 노영희, 공간정보의 이해, 국토교통부, 2015

10
• ITS : Intelligent Transportation System
https://en.wikipedia.org/wiki/Intelligent_transportation_system
Loop Detector
Traffic Monitoring RFID Reader/Traf. Sign
ITS Monitoring on GIS

11
• 360 Panorama View(StreetView)
Street View : https://www.google.com/streetview/
https://www.google.com/maps/about/partners/indoormaps/
Indoor View : https://www.google.com/streetview/#tv‐studios‐and‐sets/the‐late‐show‐with‐stephen‐colbert‐2

12
• Microsoft Photosynth
https://photosynth.net/

13
• Spatial Data Mining
https://openlayers.org/en/latest/examples/kml‐earthquakes.html

14
• 자율주행 자동차
• Google Self Driving Car
https://waymo.com/
Video : https://waymo.com/tech/

15
• Traffic Monitoring and 3D Modeling by Drone
3D : https://www.youtube.com/watch?v=SATijfXnshg
Car Detection : http://www.dlr.de/eoc/en/desktopdefault.aspx/tabid‐
5431/9230_read‐42467/

16
• 항법미사일과 공간정보기술
공간정보 융합 기술
https://topwar.ru/84614‐demony‐treh‐stihiy‐kalibr‐protiv‐tomagavka.html
토마호크 미사일 : http://kwangaeto.egloos.com/m/6250027
Tomahawk Cruise Missile
위성항법과 GIS를 이용한 순항 미사일의 유도

17
• 기상 예측과 예보
주요 공간정보 서비스
정재준, 노영희, 공간정보의 이해, 국토교통부, 2015
태풍 위성 사진 태풍의 이동 경로

19
• Spatial Data/Databases
Spatial Data Model

20
• Spatial Queries
• top‐K
• Spatial operation
• Point : kNN (k‐nearest neighbor), distance
• Line : buffer, inside, intersect
• Polygon : inside, outside
Spatial Data Model
K=4
kNN
N
distance buffer inside / intersect inside / outside

21
• R‐Tree
Spatial Data Indexing
a
b
c
d
a b c d
R
S
R S
root
root
R-tree
1
2 3
4
5
6
7
8
9
Pointers to geometries
4 5 6 7 8 91 2 3
m = 2; M = 3

22
• PostGIS/PostgreSQL : Stonebraker/Berkeley
Spatial Database Systems
Oracle
Ingres
PostgresSybaseIngres
Illustra
Postgres95
SQL Server
Relational Theory
Ingres
Informix (IUS)
1980
1970
1990
2000
PostgreSQL

23
• PostGIS
• PostgreSQL Spatial Extension
• 경부 고속도로와 Cross하는 도로를 검색하라
• SELECT gid, name FROM bc_roads
WHERE ST_Crosses( the_geom, ST_GeomFromText(‘..’, 3005) );
• 대전시내안의 도로 총 길이를 구하라
• SELECT Sum( ST_Length( the_geom ) ) / 1000 AS km_roads
FROM bc_roads;
ST_Intersects(G1,G2)
ST_Contains(G1,G2)
ST_Within(G1,G2)
ST_Touches(G1,G2)
ST_DWithin(G1,G2,D)

24
• pgRouting
• Routing function extension to PostGIS
• example page
• New York Bike Map : http://www.ridethecity.com/
SELECT * FROM pgr_dijkstra('
SELECT gid AS id,
source,
target,
cost_s / 3600 * 100 AS cost,
reverse_cost_s / 3600 * 100 AS
reverse_cost
FROM ways',
13009, 3986);

25
• New York Taxi Data
• 1.1 Billion NYC Taxi and Uber Trips
• PostGIS + pgRouting
Spatial Applications : NY Taxi
http://toddwschneider.com/posts/analyzing‐1‐1‐billion‐nyc‐taxi‐and‐uber‐trips‐with‐a‐vengeance/

26
• New York Taxi Data
Spatial Applications : NY Taxi

27
• Hadoop
• Java 기반의 대용량 데이터 처리 프레임워크
• default 3 replications : 5M files ‐> 15M files
• Hadoop HDFS :
• 구글의 2003년 논문으로 발표한 GFS(Google File System)의 구현
• 2005년에 더그 커팅(Doug Cutting)이 오픈소스로 구현
• MapReduce :
• Google의 MapReduce을 논문을 구현한 결과물
Spatial Big Data : Micro Blogs

28
• Hadoop Eco‐System
• HDFS : Hadoop Distributed File System

29
• HDFS block system : Append Only
• Default block size : 64MB
• Default replications : 3
블록1 블록2 블록3 블록4 블록5
320MB 파일
블록1
블록3
블록4
블록2
블록3
블록4
블록1
블록3
블록5
블록2
블록4
블록5
블록1
블록2
블록5
HDFS
파일저장

30
• MapReduce
Hadoop(HDFS)
MapReudce
User
Programming
map
map
map
Red
Red
Red
Red
User
Programming

31
• Spatial Social Data : Twitter
• Square Union Model
사람
지역
시간
단어Tweet
① 1차 : 2개 요소의 결합
ex) Tweet + 사람
② 2차 : 3개 요소의 결합
ex) Tweet + 사람 + 지역
③ 3차 : 4개 요소의 결합
ex) Tweet + 사람 + 지역 + 단어
④ 4차 : 5개 요소의 결합
※ 모든 질의는 top-K를 지원한
다

32
• Real‐world Social Data
• (lat, long) => key value
o1
o2
o5
o7 o4
o8
o9
o12
o10
o11
o15
o6
o13
o14
o16
o17o18
o19
o3
o20
a
b
c
e
d
f
oid wordset
o1 Italian, restaurant, expensive
o2 coffee, expensive, restaurant
o3 Italian, pizza, expensive
o4 restaurant, pizza, expensive
o5 Italian, pizza, restaurant, expensive
o6 coffee, restaurant, low-priced
o7 Italian, coffee, low-priced, restaurant, pizza
o8 coffee, restaurant, expensive
o9 expensive, restaurant
o10 pasta, pizza, expensive
o11 pasta, low-priced, restaurant
o12 Italian, restaurant, expensive
o13 pizza, low-priced
o14 tea, expensive, restaurant
o15 Italian, restaurant
o16 pasta, restaurant, expensive
o17 pizza, restaurant, low-priced
o18 Italian, pizza, restaurant
o19 Italian, pasta, restaurant, expensive
o20 pasta, expensive
(b) Social Database
(a) social data in real world
(lat, long)

33
• Simplified Tweet Data Model
• Simple MapReduce Example : Counting Tweets by User
tweet : (mid, userid, x, y, time, text)
tweet : (mid, userid, x, y, time, {word,…})
(mid, userid, x, y,
time, {word,…})
Map
(userid1, mid1)
(userid2, mid2)
(userid3, mid3)
(userid2, mid4)
(userid3, mid5)
(userid4, mid6)
shuffle Reduce
(userid1, {mid1})
(userid3, {mid3,mid5})
(userid2, {mid2,mid4})
(userid4, mid6)
(userid1, 1)
(userid3, 2)
(userid2, 2)
(userid4, 1)

34
• GeoHash : Z‐Ordering Partition
0 1 4 5 16 17 20 21
2 3 6 7 18 19 22 23
8 9 12 13 24 25 28 29
10 11 14 15 26 27 30 31
32 33 36 37 48 49 52 53
34 35 38 39 50 51 54 55
40 41 44 45 56 57 60 61
42 43 46 47 58 59 62 63
o1
o2
o5
o7 o4
o8
o9
o12
o10
o11
o15
o6
o13
o14
o16
o17o18
o19
o3
o20
(b) Spatial social data and grid space
oid wordset geo
o1 Italian, restaurant, expensive 13
o2 coffee, expensive, restaurant 18
o3 Italian, pizza, expensive 12
o4 restaurant, pizza, expensive 60
o5 Italian, pizza, restaurant, expensive 12
o6 coffee, restaurant, low-priced 35
o7 Italian, coffee, low-priced, restaurant, pizza 44
o8 coffee, restaurant, expensive 62
o9 expensive, restaurant 15
o10 pasta, pizza, expensive 12
o11 pasta, low-priced, restaurant 13
o12 Italian, restaurant, expensive 11
o13 pizza, low-priced 44
o14 tea, expensive, restaurant 15
o15 Italian, restaurant 35
o16 pasta, restaurant, expensive 15
o17 pizza, restaurant, low-priced 18
o18 Italian, pizza, restaurant 7
o19 Italian, pasta, restaurant, expensive 60
o20 pasta, expensive 62
(c) Spatial Wordset Transaction Database
Spatial
(a) social data in real world
Single Level Grid Approach

35
• GeoHash
• Z‐Ordering ID : Array of Bits
• Characterizing by Base32 Encoding
Grid Level 1
Grid Level 2
Grid Level 3
Grid Level 4
00 01
10 11
00 01
10 11
Miami, Florida의 ID? 11 11 10 = geohash : 7
Base32 Table

36
• Rowkey로서의 GeoHash
• 지오해시가 로우키를 위한 최적의 선택인 이유
① 계산하기 쉽다
② 접두사가 최근접 이웃을 발견하는데 중요한 역할을 하기 때문
• 단점 : 접두사의 정확도와 경계값 문제

37
• Spatial Aggregation : Counting Spatial‐Tweets
• GeoHash Optimization
(mid, userid, x, y,
time, {word,…}) (geoid, mid)
Map Shuffle
(geoid, {mid, mid, …})
Reduce
(geoid, n)
(‘aaaaa’, mid1)
(‘aaaab’, mid2)
(‘aaaac’, mid3)
(‘aaaad’, mid4)
(‘aaaae’, mid5)
(‘aaaba’, mid6)
(‘aaabb’, mid7)
(‘aaaa’, mid1)
(‘aaaa’, mid2)
(‘aaaa’, mid3)
(‘aaaa’, mid4)
(‘aaaa’, mid5)
(‘aaab’, mid6)
(‘aaab’, mid7)
original data Approach 1
(‘aaaa’, mid1)
(‘aaaa’, mid2)
(‘aaaa’, mid3)
(‘aaaa’, mid4)
(‘aaab’, mid6)
(‘aaab’, mid7)
Approach 2
(‘aaae’, mid5)

38
• Finding Home/Visitor
• Finding Spatial Words
(mid, userid, x, y,
time, {word,…}) (<userid,geoid>, mid)
Map Shuffle
(<userid,geoid>, {mid,
mid, …})
Reduce
(<userid,geoid>, n)
(mid, userid, x, y,
time, {word,…}) (<geoid, word>, mid)
(<geoid, word>, mid)
(<geoid, word>, mid)
Map Shuffle
(<geoid, word>,{mid,
mid, …})
Reduce
(<geoid, word>, n)

39
• SpatialHadoop
• 하둡에 공간 연산자를 지원하기 위하여 개발
• MapReduce를 이용하여 공간 연산자를 처리
• 2레벨 방식의 인덱스 이용
• Grobal index : 그리드 파일
• Local index : R‐tree
Spatial Big Data : Systems
Ahmed Eldawy, Mohamed F. Mokbel:
SpatialHadoop: A MapReduce framework for spatial data. ICDE 2015: 1352‐1363

40
• Hadoop‐GIS
• HiveSP: 공간 데이터 웨어하우징 시스템
• 공간 분할, 공간 질의 엔진을 포함
• 2레벨 인덱스 사용
Ablimit Aji, Fusheng Wang, Hoang Vo, Rubao Lee, Qiaoling Liu, Xiaodong Zhang, Joel H. Saltz:
Hadoop‐GIS: A High Performance Spatial Data Warehousing System over MapReduce. PVLDB 6(11): 1009‐1020 (2013)

41
• Spark : In‐memory Computing
• GeoSpark
• Spark의 RDD를 SpatialRDD로 변형하여 공간 연산을 지원
• 2레벨 인덱스 사용
• 지역 인덱스를 메모리에 상주 시키고 질의를 수행
Jia Yu, Jinxuan Wu, Mohamed Sarwat:
GeoSpark: a cluster computing framework for processing large‐scale spatial data. SIGSPATIAL/GIS 2015: 70:1‐70:4

42
• SpatialSpark
• GPU에서 사용하는 공간 색인 및 공간 조인을 위한
CUDA/Thrust 구현을 참고하여 개발
• Scala의 벡터/콜렉션 함수가 지원하는 병렬 기본 요소 활용
• JTS 라이브러리 사용
• 공간 조인에 대한 성능을 비교 분석할 때 사용
• LocationSpark
• Query Scheduler
• 데이터가 한쪽으로 몰려 있는 곳의 통신 오버헤드를 완화하기 위
하여 각 파티션의 통계정보를 수집하여 파티션을 다시 분할
• Query Executor
• 특정 쿼리 평가를 실행하여 질의가 인덱스를 사용해야 하는지 여
부를 판단
SpatialSpark : http://simin.me/projects/spatialspark/
LocationSpark : MingJie Tang, Yongyang Yu, Qutaibah M. Malluhi, Mourad Ouzzani, Walid G. Aref:
LocationSpark: A Distributed In‐Memory Data Management System for Big Spatial Data. PVLDB 9(13): 1565‐1568 (2016)

43
• SKSpark : Spatial Keyword Spark
• First Spatial Keyword Big Data System
• BRQ
• 사용자가 질의 영역과 키워드를 입력하면 질의 영역에 해당되는 모
든 객체 중 키워드를 모두 포함하는 객체를 검색
• ,
• = 질의 단어
• = 질의 범위
• . ∩ . ∈ ∩ . ∩ . ∈
• BkQ
• 사용자가 질의 위치와 검색 단어를 입력하면 검색 단어를 모두 포함
하는 문서들 중 사용자의 위치와 가장 가까운 순서대로 k개를 검색
• , ,
• Spatial‐Keyword In‐memory Indexing

44
• SKSpark
SpatialWebRDD Layer
GeoPartitionedRDD Layer
SKIndexedRDD Layer
Apache Spark Cluster on HDFS
Spark Layer
Query Layer
0
5
10
15
20
25
1 2 3 4 5 6
검
색
시
간
(
m
s
)
검색 키워드 수
DIR-tree
QP-tree
0
5
10
15
20
25
30
35
40
45
50
1 2 3 4 5 6
검
색
시
간
(
m
s
)
k값
DIR-
tree
BRQ
BkQ

Moving Objects
Data Computing

46
• Queries for Static Object/Static Map
• Shortest path query
• Range query : Find the restaurants
• k-NN/top-K queries
• Queries for Moving Object/Static Map
• Continuous range queries
• Location-based k-NN/top-K map queries
• Queries for Moving Objects/Moving Objects
• Internet connected cars
• k-NN/top-K moving objects queries
• Continuous moving objects queries
MO and Spatial Queries

47
MO and Spatial Queries
• Moving Objects Databases
MODB
…
MO Big Data Cluster/HDFS
MO Mining/Analysis
MO DBMS
MO Queries
Map

48
Trajectory Refinement
• Noise Filtering
• Mean(or Median) Filter
• Kalman and Particle Filter
• Heuristics‐based Outlier Detection
• v > 300km/h

49
Trajectory Refinement
• Stay Point Detection
• Semantic Filtering
• 예 : 고속도로 평균 속도/시간 계산시에 삭제/보정
• 다른 예

50
Trajectory Segmentation
• Reasons of Trajectory Segmentation
• Too large MBR : false hit in Indexing
• Semantical segmentation for mining
도심
구간
고속도로
Marios Hadjieleftherius, et. al, Efficient Indexing of Spatiotemporal Objects, EDBT 02.

51
• Methods of Trajectory Segmentation
• Time interval segmentation
• Turning points segmentation

52
• Key shape segmentation
• Stay point segmentation
Yu Zheng, Trajectory Data Mining : An Overview, ACM Trans. on Intelligent System and Technology, Sept. 2015

53
• Semantical change point segmentation

54
Trajectory Queries
• Queries
• Spatial Queries
• KAIST에 있었던 차량은?
• Spatio‐temporal Queries
• 오늘 7am‐8am 사이에 KAIST를 지나간 차량
은?
• Moving Objects Queries
• x지점 10m 이내로 지나간 차량은?
• 서로 10m 이내로 스쳐간 차량들은?
• 1시간 이상 KAIST에 머문 차량은?
• MO Mining Queries
• 두 대 이상 함께 움직인 차량들은(flock)?

55
Trajectory Queries
• Range Queries
• k‐NN Queries

56
Trajectory Queries
• k‐NN Trajectory Queries
• 예 ) Car Pool App
• Find similar trajectory cars with my trajectory

57
Moving Objects Indexing
• Indexing Moving Objects
• 3DR‐Tree for Historical Trajectory Data
• Historical trajectories are represented by their three‐dimensional
MBB(Minimum Bounding Box)
Time
Mohamed F. Mokbel, Thanaa M. Ghanem, Walid G. Aref: Spatio‐Temporal Access Methods. IEEE Data Eng. Bull. 26(2): 40‐49 (2003)
Long‐Van Nguyen‐Dinh, Walid G. Aref, Mohamed F. Mokbel: Spatio‐Temporal Access Methods: Part 2 (2003 ‐ 2010). IEEE Data Eng. Bull. 33(2): 46‐55 (2010)

58
• Indexing Moving Objects
• Multi‐version Indexing for Historical Trajectories
• Maintain an R‐tree for each time instance
• R‐tree nodes that are not changed across consecutive time instances
are linked together
Timestamp 1Timestamp 0
3D‐R‐tree
Mohamed F. Mokbel, Thanaa M. Ghanem, Walid G. Aref: Spatio‐Temporal Access Methods. IEEE Data Eng. Bull. 26(2): 40‐49 (2003)
Long‐Van Nguyen‐Dinh, Walid G. Aref, Mohamed F. Mokbel: Spatio‐Temporal Access Methods: Part 2 (2003 ‐ 2010). IEEE Data Eng. Bull. 33(2): 46‐55 (2010)

59
• TB‐Tree : Trajectory Bundle Tree
• Hybrid index structure which preserves trajectories
• A predominant trajectory index structure for Euclidean
spaces
• Fast access to the trajectory information of moving objects
Dieter Pfoser, Christian S. Jensen, Yannis Theodoridis:
Novel Approaches to the Indexing of Moving Object Trajectories. VLDB 2000: 395‐406

60
• Indexing Moving Objects for Current Position Queries
• Focus Time : Now
• Features
• Frequent Insertion of New Data
• Continuous Queries about NOW data
• Example
• A 지역에 현재 있는 차의 수는?
• 현재 x 지점 근처에 가장 가까이 있는 차량 3대는?
• Problems
• 전통적인 R‐tree는 frequent update에 느림

61
Update
Memo
Spatio‐temporal Queries
Raw answer set
Final answer set
• RUM‐tree : R‐tree with Update Memo
• Memo : R‐tree의 최근 update 포함
• update 될 때 Memo에 적어놓음
Access
Xiong, Xiaopeng, and Walid G. Aref. "R-trees with update
memos." Data Engineering, 2006. ICDE'06. Proceedings of the
22nd International Conference on. IEEE, 2006.

62
• The Time‐parameterized R‐tree (TPR‐tree)
• Minimum bounding rectangles (MBR)
• Velocity bounding rectangles (VBR)
• MBR&VBR은 차량의 속도가 유지되는 한 동일한 MBR 안에
유지 될 수 있음을 보장함으로서 update를 최소화
Simonas Saltenis, Christian S. Jensen, Scott T. Leutenegger, Mario A. López:
Indexing the Positions of Continuously Moving Objects. SIGMOD Conference 2000: 331‐342

63
Christian S. Jensen, Dalia Tiesyte, Nerius Tradisauskas:
Robust B+‐Tree‐Based Indexing of Moving Objects. MDM 2006: 12
• Bx‐Tree
• B+ tree 를 기반으로 time/spatial id를 key로 갖도록 변형
• spatial id는 Hilbert Key를 사용
• Bdual‐Tree, BBx‐tree 등으로 계속 확장

64
• Indexing on Road Network
• FNR‐Tree : Fixed Network R‐tree
• MON‐Tree : Moving Objects on Network
FNR‐Tree : E. Frentzos. Indexing objects moving on fixed networks. SSTD, 2003.
MON‐tree : Victor Teixeira de Almeida, et. al, Indexing the Trajectories of Moving Objects in Networks. GeoInformatica 9(1), 2005
R‐tree
for
Network
R‐tree
for
MO

65
Moving Objects Patterns
Fosca Giannotti, Mirco Nanni, Fabio Pinelli, Dino Pedreschi:
Trajectory pattern mining. KDD 2007: 330‐339
• Region of Interest(ROI)
• T‐Patterns
• ST Sequence Patterns
BA
20 min
C
3 min
3 min 6 min 2 min
BA
20 min
C
3 min

66
• Region of Interest(ROI) : Static ROI
• Fixed pre‐defined regions
• Find frequent visited regions
A
B
C
time

67
• Region of Interest(ROI) : Dynamic ROI
• Spatial Clustering and Labeling
• Find frequent visited regions
cluster(x1,y1)
cluster(x2,y2)
time

68
• Moving Together Patterns
• Flock
• 특정 Region r내에서 m 개체이상이 같은 방향으로 움직임
• Leadership
• 가장 먼저 움직인 객체
• Convergence
• m 개체 이상이 같은 방향으로
• Encounter
• r내에서 만나는 지역
[37] J. Gudmundsson and M.V. Kreveld. 2006. Computing longest duration flocks in
trajectory data. In Proceedings of the 14th Annual ACM International Symposium on
Advances in Geographic Information Systems. ACM, 35‐42.
[38] J. Gudmundsson, M.V. Kreveld, and B. Speckmann. 2004. Efficient detection of
motion patterns in spatio‐temporal data sets. In Proceedings of the 12th Annual
ACM International Symposium on Advances in Geographic Information Systems.
ACM, 250–257.

69
• Convoy
• min consecutive timestamps together
H. Jeung, M. Yiu, X. Zhou, C. Jensen, and H. Shen. 2008. Discovery of convoys in
trajectory databases. Proceedings of the VLDB Endowment 1, 1, 1068–1080.

70
• Swarm
• min timestamps together
Z. Li, B. Ding, J. Han, and R. Kays. 2010. Swarm: Mining relaxed temporal moving
object clusters. Proceedings of the VLDB Endowment 3, 1-2, 723–734.

71
• Trajectory Clustering
• A Partition‐and‐Group Framework
J. G. Lee, J. Han, and K. Y. Whang. 2007. Trajectory clustering: A partition‐and‐group framework.
In Proceedings of ACM SIGMOD Conference on Management of Data. ACM, 593‐604

72
• Trajectory Clustering
• A Partition‐and‐Group Framework
J. G. Lee, J. Han, and K. Y. Whang. 2007. Trajectory clustering: A partition‐and‐group
framework. In Proceedings of ACM SIGMOD Conference on Management of Data. ACM, 593‐
604

73
• Trajectory Class
• TraClass Algorithm
• Classify subtrajectories instead of whole trajectories
Jae‐Gil Lee, Jiawei Han, Xiaolei Li, and Hector Gonzalez, “TraClass: Trajectory Classification Using Hierarchical Region‐Based and
Trajectory‐Based Clustering”, Proc. 2008 Int. Conf. on Very Large Data Base (VLDB'08), Auckland, New Zealand, Aug. 2008

74
• Travel Recommendation
• Find the interesting locations and travel sequences from
trajectories
Y. Zheng, L. Zhang, Z. Ma, X. Xie, W.‐Y. Ma. 2011. Recommending friends and locations based on individual location history. ACM
Transaction on the Web 5, 1, 5‐44.
Y. Zheng, L. Zhang, X. Xie, W.‐Y. Ma. 2009. Mining interesting locations and travel sequences from GPS trajectories. In Proceedings of the
18th International Conference on World Wide Web. ACM, 791‐800.

PostTrajectory
Moving Objects Data Computing
by KNU

76
• PostTrajectory
• https://github.com/awarematics/posttrajectory
PostTrajectory Architecture
76
PostgreSQL DB
PostGIS PostTrajectory
Trajectory
Type
Trajectory
Indexing
PostGeoPhoto
GeoPhoto
Type
GeoPhoto
Indexing
Geometry
Type
R-Tree on
GiST
BlackBox App for Cars
(SkyBox)
Apache/Tomcat Web Sever
PostGeoVideo
GeoVideo
Type
GeoPhoto
Indexing
Trajectory Storage Trajectory Queries Trajectory Analysis Trajectory Apps
GeoV/P Storage GeoV/P Queries GeoV/P Analysis GeoV/P Apps
Trajectory/GeoPhoto/GeoVideo
Web User
Geospatial Media Contents Management System(GeoCMS)

77
• Data Type
• TPoint and TDouble
• TPoint : (Point, Timestamp)
• TDouble : (Double, Timestamp)
• TBool : (Boolean, Timestamp )
• MPoint
• TPoint[]
• MDouble
• TDouble[]
• MBool
• TBool[]
Data Model

78
• Data Type and Operations
Data Model
t
x
distance : moving double
MDouble
MPoint
MBool : is Intersect?
false[t1], false[t2], true[t3) …

79
• Spatiotemporal Relationship Operations
• Enters
• Leaves
• Crosses
• StayIn
• Bypass
Data Model
cross
leave
stayinenter bypass
t
x
polygon r
x
x
x

80
• Slice Operations
• Temporal Slicing
• Spatial Slicing
• Spatiotemporal Slicing
Data Model
t
polygon r
temporal
slicing
spatial
slicing
spatiotemporal
slicing
mpoint slice(mpoint, period)
mpoint slice(mpoint, geometry)
mpoint slice(mpoint, geometry, period)
mpoint slice(mpoint, period)
mpoint slice(mpoint, geometry)
mpoint slice(mpoint, geometry, period)

81
• Projection Operations
• Spatial Projection
• Temporal Projection
Data Model
LineString sproject( mpoint )
periodtproject( mpoint)
LineString sproject( mpoint )
periodtproject( mpoint)
temporal
projection
spatial projection
t
LineString
start timestamp
end timestamp

82
• Table Creation
Storage Model
CREATE TABLE taxi (
taxi_id int,
taxi_number char(20),
taxi_model char(20),
taxi_driver char(20)
traj MPOINT
);
SELECT AddTrajectoryColumn('public', 'taxi', 'traj', 4326, ‘MPOINT', 2, 150);

83
• Tables
• Meta Data Table : Trajectory_Column
• Moving Objects Table
• Segment Table
• A trajectory is split into segments
Storage Model
PostGISPostGIS
spatial objects table(roads)
PostTrajectoryPostTrajectory
trajectory_columns(metadata)
geometry_columns(metadata)
spatial objects table(usstates)
moving objects table(taxi)
moving objects table(bus)
segment table(mpoint)
segment table(mpoint)

84
• MPoint
• (segtable_oid, moid)
• One segment table per a trajectory attribute
Storage Model
PostGISPostGIS
,,,
spatial objects table(usstates)
geometry
geometry
geometry
geometry
geometry
geometry on a row
PostTrajectoryPostTrajectory
,,, mpoint
segment
segment
segment
…
mpoint := (segtable_oid, moid)
moving objects table(taxi)
segment table
moid
moid
moid
,,, mpoint
segmentmoid

85
• Example
Storage Model
CREATE TABLE taxi (
taxi_id int,
taxi_number char(20),
taxi_model char(20),
taxi_driver char(20)
);
SELECT AddTrajectoryColumn('public', 'taxi', 'traj', 4326, ‘MPOINT', 2, 150);
SELECT AddTrajectoryColumn('public', 'taxi', ‘accel1', 4326, ‘MDouble', 2, 150);
georeference
WGS4326
2 Dminsion
150 tpoint
per a segment
taxi_id traj accel1
taxi table
…
traj_segment_table
accel1_segment_table

86
• Segment Table
• strategies
• split
• count‐based‐split(*)
• spatial‐based‐split
• temporal‐based‐split
• st‐based‐split
• compression
• no compression : tpoint[]
• naïve compression : zip
• simplification
Storage Model
moid segid tpcount rect start_time end_time tpseg
box2d timestamp tpoint[]

87
• Insert new Moving Object
• Append GPS Trajectory for a moving object
• Remove GPS Trajectories
Trajectory Queries
##Inserting Moving Objects
insert into taxi values(1, '57NU2001', 'Optima', 'hongkd7');
insert into taxi values(2, '57NU2002', 'SonataYF', 'hongkd7');
UPDATE taxi
SET traj = append(traj, tpoint(st_point(200, 200),
timestamp '2010‐01‐25 12:05:30+09'))
WHERE taxi_id = 1;
UPDATE taxi
SET traj = remove(traj, to_timestamp(12345678), to_timestamp(23456789) )
WHERE taxi_id = 1;

88
• Temporal Slicing
• Spatial Slicing
Trajectory Queries
SELECT tj_slice( traj,
timestamp '2010‐01‐26 14:50:40+09',
timestamp '2010‐01‐26 15:20:40+09')
FROM taxi;
SELECT tj_slice(traj,
geometry(‘polygon ( ( 300 200, 300 300, 440 300, 440 200, 300 200 ) )'))
FROM taxi;

89
• Composite Query with Slicing
Trajectory Queries
SELECT tj_slice( traj,
timestamp '2010‐01‐26 14:50:40+09',
timestamp '2010‐01‐26 15:20:40+09')
FROM taxi
WHERE tj_overlap( traj, tj_period( to_timestamp(2432432343), to_timestamp(2432433000));
SELECT tj_slice( traj, timestamp '2010‐01‐26 14:50:40+09', timestamp '2010‐01‐26 15:20:40+09')
FROM taxi
WHERE
tj_overlap( tj_slice(traj, geometry(‘polygon ( ( 300 200, 300 300, 440 300, 440 200, 300 200 ) )')),
tj_period(timestamp '2010‐01‐26 15:00:00+09', timestamp '2010‐01‐27 00:00:00+09'));

90
• Spatiotemporal Predicate
• Spatiotemporal Predicate with Slicing
Trajectory Queries
SELECT count(*)
FROM taxi
WHERE tj_enter(traj, geometry(‘polygon ( ( 300 200, 300 300, 440 300, 440 200, 300 200 ) )'))
SELECT taxi_id, tj_slice(traj, geometry(‘polygon( ( 300 200, 300 300, 440 300, 440 200, 300 200 ) )‘))
FROM taxi
WHERE tj_enter(traj, geometry(‘polygon ( ( 300 200, 300 300, 440 300, 440 200, 300 200 ) )'))

91
• Simple Distance Queries
• Distance in WHERE
Trajectory Queries
SELECT taxi_id, tj_distance(traj, geometry('Point( 50 50 )' ),
tj_mindistance(traj, geometry('Point( 50 50 )' ),
tj_maxdistance(traj, geometry('Point( 50 50 )' )
FROM taxi;
SELECT taxi_id, taxi_number
FROM taxi
WHERE tj_mindistance(traj, geometry(‘point( 50 50 )') < 20;

92
• Join Distance
• Indexing : Query Materialization Approach
Trajectory Queries
SELECT taxi_id, bus_id, tj_distance( t.traj, b.traj)
FROM taxi t, bus b;

93
• Trajectory Data : T‐Drive
• Microsoft Research Asia, Beijin
• 6‐month real dataset of 30,000 taxis in Beijing
• Total distance: almost 0.5 billion (446 million) KM
• Number of GPS points: almost 1 billion (855 million)
Performance Evaluation

94
• Increasing Number of Moving Objects

95
• Increasing Query Window Size
• for 500k moving objects

PostGeoMedia
GeoPhoto Data Computing
by KNU
96

97
• New Era of Spatial Data/Information
• GPS Everywhere!
Introduction
images from
Apple
https://www.wired.com/2009/06/black‐box‐for‐the‐car/
http://www.bayareabikeshare.com
CarPhone Bike
and, life log for human

98
• Real Example : GeoPhoto
Introduction
JPG : EXIF(EXchangable Image File format)

GeoMedia : GeoPhoto
GeoPhoto
GPS Tagged Photo
(x, y, z)
Geophoto
KML Camera, Lookat + Field of View
Geophoto  GeoPhoto
Field of View(FoV)

100
• Real Example : Aerial Image and Trajectories in Drone
GeoMedia : GeoPhoto

101
Related Work : Flickr
이 지역의 사진을
검색하고자 한다면?

102
• Example : Finding GPS Photos in Flickr
지역/시/주/우편번호 입력지역/시/주/우편번호 입력
컨텐츠의 지정 : face, bike, building, …

103
• Example : Finding ‘Face’ Pictures in Minneapolis

104
• Example : Finding ‘bike’ Pictures in Minneapolis

105
• Current : Contents‐based Services on Photos/Videos
• Example : Face Detection using OpenCV
Related Work : OpenCV
OpenCV
Open
Source
Computer
Vision Library
http://opencv.org/

106
• Example : TensorFlow Results
Related Work : TensorFlow
https://www.tensorflow.org/tutorials/

107
• Contents‐based GPS Media Services
Traditional Approach
PostgreSQL/PostGIS
Metadata
Photos
Media ProcessingMedia Processing
Geo Media Processing
OpenCV/
TensorFlow
OpenCV/
TensorFlow
Apache/Tomcat

108
• Solution : PostGeoMedia
• PostgreSQL/PostGIS + OpenCV/Tensorflow
Our Approach
PostgreSQL/PostGIS with Media Processing
Metadata
Photos
OpenCV/
TensorFlow
OpenCV/
TensorFlow
Geo Media Processing
Apache/Tomcat

109
• Architecture
System Overview
PostgreSQL
PostGIS
PostGeoMedia
Data Types
Functions
PostGeoMedia
GeoMedia
Data Types
Function_1
Function_2
Function_3….
Index
PL/pgsql
PL/python OpenCV
EXIF
Tool
Tensor
Flow

110
• Creation of GeoPhoto Tables
GeoPhoto Queries
CREATE Table dronePhotos
{
ID int,
owner VARCHAR(20),
img GeoPhotos
}
Select AddGeoPhotoColumns( photos, ‘img’ );
CREATE Type GeoPhotos
(
ID          INT,
uriString VARCHAR(100),
width       double,
height      double,
time        TIMESTAMP,
geom GEOMETRY(POINT,4326)
);
INSERT INTO dronephotos VALUES( 1, ‘kwnam’, 'file://tmp/drone001.jpg');

111
• Face and Eye Detection in SQL
• Spatial Queries with Face/Eye Detection
GeoPhoto Queries
SELECT faces( uriString), eyes(uriString), cars( uriString )
FROM DronePhotos;
SSELECT faces( uriString ), eyes( uriString), cars( uriString )
FROM DronePhotos
WHERE ST_WITHIN( geom, ST_MakeEnvelope(191232, 243117,191232, 243119,312) );

112
• Car Detection in Drone Images
GeoPhoto Queries
Car Detection : http://www.dlr.de/eoc/en/desktopdefault.aspx/tabid‐
5431/9230_read‐42467/

113
• 사람이 찍힌 사진을 검색하시오
• 특정 지역내에 사람이 찍힌 사진을 검색하시오
GeoPhoto Queries
SELECT count(*)
FROM DronePhotos
WHERE NumOfFaces(uriString) >0 OR NumOfCars( uriString ) > 0 ;
SELECT count(*)
FROM DronePhotos
WHERE ( NumOfFaces(uriString) >0 OR NumOfCars( uriString ) > 0 ) AND
ST_WITHIN( geom, ST_MakeEnvelope(191232, 243117,191232, 243119,312) );

114
• Join
GeoPhoto Queries
SELECT count(*)
FROM DronePhotos dp, ROAD r
WHERE ST_WITHIN( dp.geom, ST_BUFFER( r.geom, 10 ) );

115
• Privacy Protection or Enhancement
GeoPhoto Queries
SELECT myblur( faces(uriString) )
FROM photos
WHERE ST_WITHIN( geom, $USA );
SELECT myenhance( faces(uriString) )
FROM photos
WHERE ST_WITHIN( geom, $USA );

116
• 번호판 인식 + GPS
Applications
http://blog.naver.com/emaru2017/221070385130

118
Example
• Geospatial Augmented Photo

PostGeoMedia
GeoVideo Data Computing
by KNU

120
• Real Example : GeoVideo/Trajectories in Car Blackbox
• Moving Point and Moving Double
GeoVideo
*.mp4 : video data
*.gps : gps data(NMEA)
*.3gf : acceleration sensor

GeoVideo : Moving Objects
• GPS Trajectory and GeoVideo의 분석과 모니터링
차량 3대
센서 데이터
비교 분석

122
• MediaQ by USC
• GeoUGV : User‐Generated Mobile Video
• Since 2014
• MediaQ web: http://mediaq.usc.edu
Related Work
Ying Lu, Hien To, Abdullah Alfarrarjeh, Seon Ho Kim, Yifang Yin, Roger Zimmermann, and Cyrus Shahabi,GeoUGV: User‐Generated
Mobile Video Dataset with Fine Granularity Spatial Metadata, In the 7th ACM Multimedia Systems Conference (MMSys), Klagenfurt
am Worthersee, Austria, May 10‐13, 2016
Seon Ho Kim, Ying Lu, Giorgos Constantinou, Cyrus Shahabi, Guanfeng Wang, and Roger Zimmermann,MediaQ: Mobile Multimedia
Management System. In the conference of Multimedia Systems (MMSys), In the conference of Multimedia Systems (MMSys), pages:
224 ‐ 235, Singapore, March 19 ‐21, 2014

123
• GeoVideoIndex by POSTECH
• GeoVideo
• MBTR : Minimum Bounding Tilted Rectangle
Related Work
Dongha Lee, Jinoh Oh, Woong-Kee Loh, Hwanjo Yu:
GeoVideoIndex: Indexing for georeferenced videos. Inf. Sci. 374: 210-223 (2016)

PostGeomedia
GPS VideoSensor
Trajectory
SubTable
Blackbox Drone
index
Moving Feature Table
Query UDF Extension
PostGeoMedia Integration Layer
Sensor
SubTable
index
Video
SubTable
index
• 공간 미디어 인덱스
Indexes :
3DR Tree on GiST
SETI* on ORDB
Bx* Tree on GiST
Storage :
Traditional Table
Plain Binary Objects
Compressed WKB
Materialized
View

125
Conclusion
• Trends
• Spatial Big Data Computing
• Moving Objects Data Computing
• Our Systems
• SKSpark
• PostTrajectory
• PostGeoMedia

Moving Objects and Spatial Data Computing

Recommended

Recommended

More Related Content

Similar to Moving Objects and Spatial Data Computing

Similar to Moving Objects and Spatial Data Computing (20)

More from Kwang Woo NAM

More from Kwang Woo NAM (20)

Recently uploaded

Recently uploaded (20)

Moving Objects and Spatial Data Computing