Non-intrusive velocimetry techniques using image processing techniques

Size: px

Start display at page:

Download "Non-intrusive velocimetry techniques using image processing techniques"

笈幸奚
5 years ago
Views:

1 m (m) <10 20~ ~ < () (m) <50 50~ ~ ~ ~ < ex. 100m (m) 0.7~ ~ ~ < (m)

2 ADCP Tracer Dilusion ADCP 150 3m/s BURNELLIESTIMATION OF VELOCITY PROFILES IN NATURAL CHANNELS DURING HIGH FLOODS

course webpage: https://riverflow2018.irstea.

3 Stream Gauging Course Almost all the presentations are now available on the training course webpage: UAV etc., ~

4 CCTV UAV :,, etc. HP 30 QH :Qf(H)

5 DVRTB m m JR 4 939m R m 2 480m R m

6 #1 UHF 2 #2 UAV() avi, mpeg, mp4, MTS, AVCHD, h fps (frame per second) NTSC 25 fps PAL 4K : 3840 by 2160 pixel HD : 1920 by 1080 pixel VGA : 640 by 480 pixel QVGA: 320 by 240 pixel

Float PTV : Particle Tracking Velocimetry LSPIV : Large-Scale Particle Image Velocimetry PIV Aerial LSPIV LSPIV STIV : Space-Time Image Velocimetry Aerial STIV #1.

Use of High-accurate clinometer #8. Use of video images from moving helicopter All of the above techniques are developed at Fujita Lab. Kobe Univ.

7 Float PTV : Particle Tracking Velocimetry LSPIV : Large-Scale Particle Image Velocimetry PIV Aerial LSPIV LSPIV STIV : Space-Time Image Velocimetry Aerial STIV #1. Large scale Particle Image Velocimetry (LSPIV) #2. Aerial LSPIV #3. Space Time Image Velocimetry (STIV) #4. Measurement by STIV at night #5. Aerial STIV #6. Space Time Volume Velocimetry (STVV) #7. Use of High-accurate clinometer #8. Use of video images from moving helicopter All of the above techniques are developed at Fujita Lab. Kobe Univ. Japan X,Y,Z(x,y) A(X i,y i,z i ) a GCP (a) Physical coordinates (X,Y,Z) Video camera X Y Z y x flow A(x i,y i ) a (b) Screen coordinates (x,y) DLTDirect Linear Transformation) A1 X A2Y A3 Z A4 x C X C Y C Z 1 1 Colinearity relations 2 3 B1 X B2Y B3Z B4 y C X C Y C Z 1 a X X a Y Y a Z Z 11( 0) 12( 0) 13( 0) x c a31( X X 0) a32( Y Y0 ) a33( Z Z0) a X X a Y Y a Z Z 21( 0) 22( 0) 23( 0) y c a31( X X 0) a32( Y Y0 ) a33( Z Z0) 1 2 x o y o 3 (x 0,y 0 ) : (X 0, Y 0, Z 0 ) : c: a ij : 10cmm Z D X 1 D2Y D3 DLT

8 Float PTV m LSPIV PIV template) a(i, j) b(i, j) a,b a(i, j) b(i, j) Search area a(i, j) b(i, j). STIV STIV STIV Space-Time Image 100 t=0 s t=1/30s t=2/30s 300 t = 300/30 =10s

9 STIV STIV STIV (C) pixel300pixel (16.2m10sec) 257pixel300pixel (20.5m10sec) 184pixel300pixel (15.2m10sec) 16.3m 19.0m (Deg), J pq A g g dxdt x x p q 2J12 tan 2 J J sec 10sec A() B() C() N i1 ( i) C( i) N i1 C( i) 3.2 m/s 3.1 m/s 53

10 m 17.7m Q) V m V surface Surface Velocity Coefficient Q i i xv mi 5sec 56 ISO748 u( z) u * 1 ln z z 0 u b * ghs 0.41 k s z0 10 u( z) V Surface z h 1/ m m m 6~ ~3 U*:Sks log V V m surface h h z 0 ln z z 0 1 power V V m surface m m

11 STIV STIV, LSPIV 0.85 H-ADCP FIR camera vs HD camera C Date April 19~22,2012 downstream of Br. (Camera AFIR) (Camera BCHD) floatadcpradio wave FIR camera FLIR pixel 7.5Hz pixel 30Hz m2ts A B D Negoya Br. HD camerasony 640x x540 E flow April

12 the Uono River, Negoya Bridge the Uono River, Negoya Bridge the Uono River, Negoya Bridge. : HD camera : FIR camera the Uono River, Negoya Bridge. April 21, 10:00 =17.8m = 5.6m x 25 = 140m April 21, 10:00 =17.8m = 5.6m x 25 = 140m

13 ADCP :00 : STIV vs ADCP FIR camera HD camera ;20:00- STIV vs ADCP vs float( ( ;19:00-

14 STIV 1 2 Generate STI for a search line Water level variation can be captured by image analysis Seach line # Under water 83 84

85 STIV2014330 Application of STIV(March,30,2014 9:10 3.23m/s 4.44m/s 3.

13500 9:45 14400 10:00 Mark points length4.65m 6.8 5 6 4.5 5.2 4 4.4 3.5 3.6 3 2.8 2.

2 100% 50% 30% 0% 0 7200 8:00 8400 8:20 9600 8:40 10800 9:00 12000 9:20 13200 9:40 14400 10:00 m 3 /s 50 45

15 85 STIV Application of STIV(March,30,2014 9: m/s 4.44m/s 3.96m/s s time10 STIV(March,30,2014 (m 3 /s) : : : : : :00 Mark points length4.65m Surface velocity distribution 86 m % 50% 30% 0% : : : : : : :00 m 3 /s % 50% 30% 0% :00 10: : : : : : (m) Aerial LSPIV Aerial LSPIV: 2007 Aerial LSPIV: 2007 #

16 Aerial LSPIV: 2007 Aerial LSPIV: 2007 Fast,large-scale,particleimagevelocimetry-basedestimationsofriver surface velocity DavidW.Dobson a,n, K.ToddHolland b,1, JosephCalantoni Computers &Geosciences70(2014)35 43 Performance of image-based velocimetry (LSPIV) applied to flash-flood discharge measurements in Mediterranean rivers J. Le Coz a,*, A. Hauet b, G. Pierrefeu c, G. Dramais a, B. Camenen Journal of Hydrology2010 Performance of image-based velocimetry (LSPIV) applied to flash-flood discharge measurements in Mediterranean rivers 283 UAV 110 SfM / MVSStructure from Motion / Multi-View Stereo

17 River Flow 2014 Helicopter-based surface PIV experiments at Thur River M. Detert & V. Weitbrecht Laboratory of Hydraulics, Hydrology and Glaciology VAW, ETH Zurich, Zürich, Switzerland Figure 7 Time-averaged surface velocity field with raster resolution of m2 and 50% overlap, i.e. vector spacing of 0.5 m, and streamlines UAV -> KU-STIV ()

18 FLOW 2012 STIV 2013 ADCP 2014 Aerial STIV ADCP ADCP ADCP 300m m10sec 23.4m10sec STI No.15 STI No STI No.5 KU-STIV Aerial-STIV STI No.15 KU-STIV Aerial-STIV Aerial STIV ADCP ADCP U (m/s) U (m/s) H (m)

19 STIV : :

20 15:40 NHK No.3 No.6 No.13 No.3: 4.58m/s No.6: 5.25m/s No.13: 4.56m/s 0.86 s Blender Marker Tracking LP18.29m pixel(0.06m/pixel) 15.08m m m

21 (02:10) (02:57) (03:16) (04:46) levee breach width (m) :00 13:00 14:00 15:00 16:00 17:00 18:00 15: time 17:5 7 1 inundated water volume (m 3 ) (14:00) (15:00) (16:00) (17:00) elapsed time (min) STIV 12:50~16: *10(m³) 0.78*10(m³)41% m/582 pixel PIV 40x40(2.4x2.4m) 13:5726 (0.87s) UAV Space Time Volume Velocimetry (STVV) 1. In stead of generating space-time image (STI), space-time volume (STV) is generated in STVV.

22 STVV LSPIV 1. Image transformation can be executed simply by using - focal length - height between lens and water level - sensor size of CCD or CMOS Total length=858m ASP 1. GCP 2. (ASP) 3. LSPIV Original image shot from riverbank Ortho-rectified image LSPIV KU-STIV FFTSTI STI

23 KU-STIV Y X KU-STIV

Microsoft Word - °ò·þ®{ÄÝÆFªº¸g¾ú 4.doc

Microsoft Word - °ò·þ®{ÄÝÆFªº¸g¾ú 4.doc 基督徒屬靈的經歷第四課 : 靈泉出埃及記十五章 22 節到 27 節 : 摩西領以色列人從紅海往前行, 到了書珥的曠野, 在曠野走了三天, 找不著水到了瑪拉, 不能喝那裏的水, 因爲水苦, 所以那地名叫瑪拉百姓就向摩西發怨言, 說, 我們喝什麽