The Argus Runtime Environment Guidelines on
Contents
1. Important note Notice that all Argus processing is performed in GMT times For instance geometries are selected on the basis of their whenValid time which is in GMT and fielddata need to be defined against GMT time Local times show up in only 2 situations of relatively minor importance 1 in the non used readable part of the image filename see below and 2 in the whenDone field of the geometry solution 2 3 Site user amp platform dependent settings Site user amp platform dependent settings are specified in the file argusOpt m Reference is made to Appendix A for a detailed description of the contents of argusOpt manualArgusTools2002b 5 Version dd 28 11 02 3 The Argus Runtime Environment ARE This chapter provides further background on the components of the ARE Section B 2 as well as an introduction on the use of the argusTools Section B 3 In anticipation of that Section B 1 discusses the initialization of the ARE 3 1 Initialization of the ARE In anticipation of the analysis of Argus video data the ARE needs to be initialized Initialization of the ARE involves two aspects 1 Identification of the Argus meta database of interest This implies that the ARE reads the appropriate station camera gcp geometry etc information from file 2 Inclusion of all appropriate Matlab paths to the directories which contain the argusTools source code Running the function argusInit which is located in t2. default settings for the plan view area panoramic field of view and color scaling are read from file argusMergeTool gui getSiteSpecifics and filled in the upper right section of the database Notice that variation of the spatial DX DY and angular Dtheta Dalpha resolution directly affects the resulting image size To remove sharp transitions between different cameras an optional colour balancing can be applied The method presently implemented is based on the work done by Jack Puleo Naval Research Lab MS USA It scales image intensities per colour band on the basis of a mean Rmean Gmean Bmean and offset value Rmin Gmin Bmin A side effect of the use of colour balancing may be that the image overall looses contrast Further investigation over a range of Argus sites is needed to arrive at optimum settings for the colour balancing parameters The lower section of the AMT user interface governs which images are to be merged Merging can be performed in single image mode output to screen in bulk mode output stored on disk or from file In the latter case the user needs to create an input file text format which provides a listing of the images to be merged Example input files are provided in Appendix B In the case bulk merging an arbitrary number of images per day can be merged during a longer period of time In this specific case the user can specify e The type of output he prefers jpg images or mat files or both e Whet
3. separate window and the computed GCP s are shown in the image w ma manualArgusTools2002b 10 Version dd 28 11 02 12 Modify the whenValid field If not the time in the whenValid field will be set to match the epochtime of the image you are presently working on However the situation may occur that the new solution was valid already during a certain time ahead of the present image e g during the morning of the same day In that case response with Yes Geomtool2002 will invite to select an Argus image starting from which the new geometry solution will be valid 13 Judge on the quality of the result by comparing the computed location of the GCP s to their real world location as observed from the underlying image If you are not satisfied yet re determine a geometry solution by excluding some picked GCP s from the analysis or by picking additional GCP s 14 If you are happy with the result save the new geometry to file menu geometry save geom 15 Refresh the image menu image refresh and remove all picked GCP s manually button select DB picked GCP s clear UV for each of them Load the next image 16 Repeat steps 2 15 Notice that geomtool2002 offers some additional functionality which is worth mentioning here e Option rectification menu Image Plan view This option invites the user to indicate the 4 corners of an area of interest by clicking the left mouse button After the fourth point geom
4. Argus users are invited to add their own core routines to this directory e Two local settings files named argusInit m and argusOpt m These files are needed to initalize the ARE and to define the location user platform settings During the installation of your ARE both need manual editing see below The two files are located at the same level as all directories mentioned above For reasons of uniformity it is recommended to leave them there Besides the directories mentioned here your ARE may contain a directory argusDB msql This directory is only relevant for use of the ARE at Oregon State University Setting your Argus Runtime Environment The ARE is installed by copying the rootdirectory argus to your local harddisk Before being able to use your ARE the Matlab files argusInit and argusOpt need to be edit manually For both files the lines that may need manual editing are located in a separate section immediately at the beginning of the file manualArgusTools2002b 21 Version dd 28 11 02 In the file argusInit m the name of your local argusDB is defined including the path The following restrictions apply e Your local argusDB has to be a subdirectory of the directory argusDB matlab at the same level as the databases delft osu plymouth etc e The name of your argusDB has to match the name of the directory cf delft osu etc In the file argusOpt m the following local s
5. file contains the time information that is needed to specify the merge of interest Preceeding comment lines are ignored Time information can be defined in 3 different ways 1 yyyy mm dd GMThr 2 yyyy mm dd GMThr minute 3 Argus image filename The example file listed below combines the 3 types of time definition Input file for argusMergeTool AMT IMPORTANT NOTE IF TIMES OF INTEREST ARE DEFINED THROUGH THE NAMES OF ARGUS IMAGES OPTION 3 ONLY THE EPOCHTIME NUMBER IS USED SO NO ACCOUNT IS TAKEN OF THE CAMERA NUMBER IMAGE TYPE ETC THESE ARE DETERMINED FROM THE SETTINGS OF THE UPPER PART OF THE GUI AP dP dP P dP AP AP dP 1999 05 31 10 1999 05 31 11 1999 05 31 12 20 1999 05 31 14 928138784 Mon May 31_08_19 44 GMT 1999 jvspeijk cl timex jpg 928142384 Mon May 31_09 19 44 GMT 1999 jvspeijk cl timex jpg 928145985 Mon May 31_10_19_45 GMT 1999 jvspeijk cl timex jpg manualArgusTools2002b 24 Version dd 28 11 02 Appendix C DEM on the basis of shorelines This appendix provides an overview of the Matlab commands which allow for the set up of a Digital Elevation Model or Mapping the intertidal beach bathymetry on the basis of video derived shorelines This type of post processing is not a standard feature yet within the ARE Before running these commands make sure that the shorelines of interest as well as the support file breakwater mat not crucial are located in the Matlab s working directory Otherwis
6. interest get wave conditions on the basis of an Argus image filename epochtime of interest or period of interest get meteorological conditions on the basis of an Argus image filename epochtime of interest or period of interest Table 4 4 Overview of Argus routines for including field data Both DBGetEuroTide m DBGetEuroWave m and DBGetEuroMet m account for 1 preference and 2 preference data sources 4 5 Argus pixel processing The collection and analysis of pixel time series and time stack images presently attracks considerable attention within the Argus Users Group The code and tools developed to that end by Oregon State University are in the user driven stage Background information on the use of the Argus PIXel Toolbox is provided in a note by John Stanley and Rob Holman which is included as Appedix D to this manual manualArgusTools2002b 19 Version dd 28 11 02 5 References Aarninkhof S G J and Roelvink J A 1999 Argus based monitoring of intertidal beach morphodynamics Proc of Coastal Sediments Conf Long Island NY USA 2429 2444 Aarninkhof S G J Caljouw M and Stive M J F 2000 Video based quantitative assessment of inter tidal beach variability Proc of Int Conf on Coastal Engineering Sydney Australia Enckevort I M J and Ruessink B G 2001 Effect of hydrodynamics and bathymetry on video estimates of nearshore sandbar position Journal of Geophysical Research 106 C8 16969 16
7. was extended to operate on other database elements as well sites stations cameras amp GCP s Allthough DBO allows for the modification of each of these structures it is strongly recommended not to change any field of the site station camera or GCP structure since this may strongly violate the functioning of the ARE hag Dare Flaws Daran ARG CGC Zu Se Came Ea eee Emori Jian van Sek Lighthouse Comer Whee hed Lenstoe a Ei Are RETTEN Gul ahearlices Hia i ee GHT Lew rrr dee LTE Tore Bat Yale Geometry Schr tes EH Sea IE Foy Tel i 7 0 N T Pr zu 120 NIE 1 Dee AU OLS seein 1 I TE Tle E se 1 12 PIE Pk CL LEN PAT PA Rese CT aL Ara ot Salen 1d FUT Fee emi al 198 TEE 0 Map AIN D ye uE TAU ANT UE barr AE DI I 3 T ure a LLAN it Duaiiy 1 5 tha rh 4 ee ee Coma I ir hay hong pin 12 AE DA Sen D el Duni arm ama KE rT 1 Foire CH aoe Ch i ira BSR He Ca sl 1m 1m3 w 2a a gt So Ze ee Zeil Fig 3 1 User Interface ofthe Argus DataBase Organizer DBO The relevant functionality of DBO can be described the next listing 1 Make sure the ARE has been initialized Start DBO by typing DBO at the Matlab prompt DBO automatically loads the entire current Argus database that is sites stations geometries eurotides etc 3 View the information available You can swap between data elements with the help of the Select DB button in the lower left corner Some of the elements provide sorting functionality to f
8. 2LocalString m Converts epochtime to a string definition wrt local time julian2Matlab m Converts Julian day number to Matlab s datenum format Accounts for leap year cycle Accounts for leap year cycle argusDay m Routine to compose Julian day from epoch time or Argus struct with field time Overlapping with matlab2Julian m Table 4 1 Overview of Argus routines for time processing Associated Matlab functions of interest now date datenum datestr datevec 4 2 Accessing argusDB information Table 4 2 provides an overview of the most important routines to retrieve Argus meta information from the argusDB manualArgusTools2002b 16 Version dd 28 11 02 Application Connect to argusDB msql only Get name of argusDB presently in use DBCreateEmptyStruct m Create empty argusDB structure DBGetAllGeometries m Get all geometries for a specific station shortName BT like nordzeel and camera number Returns all sites of argusDB presently in use DBGetCameraByID m Get camera structure given cameralD eg NOO1C Benz Output includes meta information on station amp lens DBGetCamerasByStation m Return data about cameras for a specific stationID at a pee gee DBGetCurrentGeom m Return the most recent geometry solution for a specific een Sane eg Noo ata ghene o DBGetGCPBySiteID m Return all GCP s at a site for a specific siteID eg e YAQUINA at a given epochtime DBGetImageData m Returns Argus meta inf
9. 980 Holland K T R A Holman T C Lippmann J Stanley and N G Plant 1997 Practical use of video imagery in nearshore oceanographic field studies IEEE Journal of oceanic engineering Vol 22 No 1 Lippmann T C and R A Holman Quantification of sand bar morphology A video technique based on wave dissipation Journal of Geophysical Research 94 C1 995 1011 1989 manualArgusTools2002b 20 Version dd 28 11 02 Appendix A Setting up your Argus Runtime Environment This appendix discusses some issues which are of relevance for setting up your personal Argus Runtime Environment ARE Argus Runtime Environment Directory structure amp contents Supply of the ARE by WL delft hydraulics involves the provision of licensed Argus analysis software on a cd or through the Internet The supply covers all components of the ARE except for the Argus image archives which are assumed to be available already However a subset of Argus images sampled at different sites world wide can be delivered upon request The ARE is supplied in a single directory named argus which has the following contents e A directory argusDB which contains all core routines that are used to access the Argus database of meta information These routines are supposed to be used by advanced Argus programmers e A directory argusDB matlab which contains the Argus databases of meta information in Matlab format The databases are organized i
10. At the same time all GCP s available at the site and time of interest are imported Label the GCP s on the screen menu image show labels 4 Choose a clearly visible GCP which can be used to determine a new geometry solution Zoom in somewhat within the area of this GCP click on left mouse button 5 Select the GCP of interest from the listing of GCP s make sure the lower left button is set to All GCPs Pick the location of the GCP on the screen menu GCP Pick gcp Zoom out again right mouse button 8 Repeat steps 4 7 for the next GCP You need at least 2 to be able to determine a geometry solution 9 A very clear horizon can be used as a GCP as well To that end select GCP horizon gcp from the menu and follow the instructions click just a few pixels above the horizon use left mouse button finish by clicking the right button Decide whether you accept all horizon gcp s 10 If any of the GCP s does not satisfy than select the option Picked GCPs from the selection button in the lower left corner Select the erroneous GCP and click Clear UV in the lower right corner Do so for each GCP which should not be used for the determination of a geometry solution 11 Determine the new geometry solution menu Geometry solve geom This is done by following an efficient iteration procedure If this procedure is successfully typically after about 30 80 iteration steps the result is summarized on screen
11. The Argus Runtime Environment Guidelines on Installation and Use Stefan Aarninkhof WL Delft Hydraulics Kenneth Kingston University of Plymouth Merged panoramic view Olympic Harbour Barcelona Mapfre station Spain Table of Contents 1 e processing external field Appendix A Setting up your Argus environment Appendix B Input file Argus Merge Tool Appendix C DEM on the basis of shorelines Appendix D Manual Argus PIXel Toolbox manualArgusTools2002b 2 Version dd 28 11 02 1 Introduction This note is meant for use as a starting point to explore the Argus Runtime Environment ARE Sensible and efficient analysis of Argus video images demands various data sources to come together video data Argus meta information geometry solutions etc and supporting field data see Fig 1 1 The ARE provides a framework for that by offering a set of licensed Matlab based Argus analysis tools which enable the user to quantitatively interpret video data to perform post processing on the raw data like image merging and to derive meaningful information like shorelines from the images Video images snap timex var gt Results Field data Argus data base tides waves meta information Fig 1 Layout ofthe Argus runtime environment ARE This note provides further background on each of the components of the ARE Furthermore it briefly discusses the use of the various argusTools and it gives an o
12. acilitate access to the data eg geometries can be sorted by camera whenDone and whenValid 4 Modify the data base as requested This generally concerns the removal of erroneous geometry solutions modification of the whenValid field of geometry solutions or the inclusion of new eurotide eurowave structures Notice that the date format for the whenValid field is very strict 5 Once again It is strongly recommended not to change any field of the site station camera or gcp elements 6 Save the updated database with the help of the DBO Save option 3 3 2 geomtool2002 FG Geomtool2002 was originally developed by Nathaniel Plant Naval Research Lab MS OSU making use of Argus core routines developed by John Stanley at Oregon State University The application is meant for the generation of new geometry solutions For a manualArgusTools2002b 9 Version dd 28 11 02 detailed background on image quantification and geometry solutions reference is made to Lippmann and Holman 1989 and Holland et al 1997 Fe mp IP Dam SE REES bemari i peered bere ae SeRSE Ee BE Fig 3 2 User Interface of geomtool2002 FG Use of geometool2002 typically involves the following steps 1 Load an Argus image for which the most recent geometry solution is no longer valid menu file load image 2 Load the most recent geometry solution menu geometry load geom to get a first order impression of the location of the different GCP s
13. e the dir command will fail Start of with nothing in mind amp make sure the ARE is initialized clear all close all addpath d argus argusInit pari Start date 07 Mar 2002 Now 21 Mar 2002 14701709 End date 01 May 2005 Date check OK Owner PARI License OK Argus runtime environment set up for argusDB pari Get a listing of shorelines available for September 3 2001 The result is a 10x1 structure array named fns containing amongst others the filenames created by IBM fns dir 20010903 mat fns 10x1 struct array with fields name date bytes isdir fns 1 ans name wl miyazaki 20010903 gmt0000 cl mat date T21 Mar 2002 13 45 38 bytes 23456 isdir g Collect all xyz co ordinates in one variable named xyz Notice the inner for j loop This loop is needed to account for the second and further parts of a waterline in the case that a waterline was derived from an image by multiple IBM runs that is the user applied multiple neighbouring ROI s xyz for i 1 size fns 1 load fns i name manualArgusTools2002b 25 Version dd 28 11 02 for j l length allWL xyz xyz allWL j xyz end end Define the overall DEM grid The function meshgrid is a standard Matlab routine to do so xi 150357325 yi 750 5 2400 X Y meshgrid xi yi Interpolate the raw waterline locations xyz to the DEM grid This yields an elevation matrix z
14. e of meta information DBO Data Base Organizer see below provides the functionality to do so e The ARE enables the user to specify a 1 preference as well as a 2 preference data source for each Argus site The 1 preference source of tide data is often a locally measured water level whereas the 2 preference source may be an astronomically predicted level The preference of wave data sources generally decreases with increasing distance from the Argus site of interest The preference of each data source is also stored as part of argusDB e External tide wave and meteorological data can be read from the argusDB through the routines DBGetEuroTide DBGetEuroWave and DBGetEuroMet respectively See the help on those routines for further details 3 3 The Argus Analysis Tools At present six applications are part of the standard set of Argus analysis tools These are summarized in Table 3 1 Application DBOrganizer Date Base Organizer geomtool2002 Find Geometry 6 ArgusDesignTool__ ADT Argus Design Tool Table 3 1 Overview of standard Argus Analysis Tools In the next few sub sections detailed comments amp guidelines are given on each of six applications mentioned above manualArgusTools2002b 8 Version dd 28 11 02 3 3 1 DataBase Organizer DBO Initially DBO was developed to view existing geometry solutions and if necessary to remove erroneously stored geometries from the argusDB Later on DBO
15. ettings are defined e imageArchive argusDB This variable represents a pathname which refers to the root of your local Argus image archive for a specific argusDB Notice that argusDB should exactly match with one of the databases that were defined in the routine argusInit cf delft osu etc At the level of this imageroot we find the image archives of one or more Argus stations hence directories named argus02a egmond miyazaki etc which obey a standard structure cf Section e fielddataRoot Pathname to the directory which contain the fielddata archives subdirectories Tide Wave and Meteorological as well as the Matlab code for access e timeOffset amp timeZoneString Time offset in minutes with respect to GMT of the location where the ARE user lives so Oregon 480 Japan 540 Also the associated timeZoneString like MET Middle European Time JST Japanese Standard Time etc needs to be given here e username Name of the ARE user In this way you can be credited for perfect geometry solutions After assigning appropriate values to the variables mentioned here no further modifications are needed and your ARE is ready for use Initialization of your Argus Runtime Environment Only after running the initialization file argusInit m your ARE is ready for use The function arguslnit has to be called with a single input argument character string which rep
16. he argusRoot cf Appendix A initializes the ARE Make sure that argusInit m is on your Matlabpath The function argusInit requires a single input argument which represents the name of the database of interest eg arguslnit delft argusInit plymouth etc See Appendix A for an overview of the demands regarding the naming of Argus databases 3 2 Data sources Argus Runtime Environment 3 2 1 Argus archive of video data Ideally the ARE reads the Argus video data directly from the Argus image archive If impossible for technical reasons a local copy of the image archive is to be created Notice the following aspects e The directory structure of the local image archive has to be an exact copy of the image archive structure designed by OSU This implies that the ARE expects the following directory layout imageRoot site year camera julianDay imagename for example d argusArchive egmond 1998 c 1 293_Oct 20 908878700 Tue Oct 20_10_18_20 GMT 1998 egmond c1 timex jpg e de imageRoot is to be defined in the function argusOpt In the example given above the imageRoot is d argusArchive e Generally the imageRoot is named argusImages argusArchive imageArchive etc e If the ARE is used in a unix environment all backward slashes mentioned here are to be replaced by forward slashes Or even better use Matlab s filesep command as much as possible e Image filenames are requested to have the long Argus name
17. her or not to include gridlines amp tickmarks at the merged images e The output directory This goes by selecting an existing file in the target directory manualArgusTools2002b 12 Version dd 28 11 02 NB Notice that the times specified in Section 3 of the user interface represent GMT times So requesting AMT to generate a merge of Miyazaki at September 22 GMT 06 00 hr will merge images collected at Miyazaki within half an hour of JST 15 00 hr local time 3 3 4 Argus Stack Tool AST AST was designed for the analysis of the morphodynamics of sand bars To that end image intensities are sampled along a user specified array and stacked over time On the basis of the evolution of the wave breaking patterns over time the migration of sand bars can be quantified vo ES Go Go Ge Ge na Fig 3 4 User Interface of Argus Stack Tool AST AST requests the user to specify the site of interest the end points of the array of interest and the period of interest As before the period of interest is to be specified in GMT times With the help of the option Show Array the user can verify whether the array co ordinates entered do satisfy Proper use of this application demands that image intensities are sampled at similar tidal levels cf Van Enckevort and Ruessink 2001 To that end AST applies an image filter which selects the images that obey some user specified filter criteria In the example given in Fig 3 4 on
18. i zi griddata xyz 1 xyz 2 xyz 3 X Y cubic Now replace all elements of zi which are further than 10 m away from the nearest xyz point by NaN s This is to remove interpolation induced erroneous values from the elevation matrix Notice that this operation may take a little time Nx length xi Ny length yi for x 1 Nx for y 1 Ny distance sqrt xyz 1 xi x 2 xyz 2 yi y 2 if distance gt 10 zi y x NaN end end end Load the file breakwaters mat very miyazaki specific This file contains the contours of the breakwaters in front of Miyazaki beach Just used for reference purposes in the plot load breakwaters mat Now create the plot with the help of the contourf command figure 1 subplot 311 contourf yi xi zi 0 0 25 2 manualArgusTools2002b 26 Version dd 28 11 02 Shading flat makes the result look much better Also set the axes add labels and include a colorbar shading flat set gca xdir reverse hold on plot allWL xyz 2 allWL xyz 1 k hold off axis 2400 750 100 400 set gca fontname times fontsize 12 h colorbar set h fontname times fontsize 12 q get h Title set q String Tokyo Peil set gca xticklabel ylabel x m fontname times fontsize 12 And here is the final result manualArgusTools2002b 27 Version dd 28 11 02 manualArgusToo
19. lines reference is made to Aarninkhof and Roelvink 1999 Model accuracy was investigated in Aarninkhof et al 2000 Extension of the IBM functionality by means of the inclusion of relevant post processing techniques shoreline evolution over time mapping intertidal beach bathymetry is anticipated in the future manualArgusTools2002b 15 Version dd 28 11 02 4 Advanced Argus programming Chapter 4 provides an overview of Argus core routines that can be applied for advanced Argus programming It strongly recommended to use these routines as much as possible since they are well tested it strongly facilitates the exchange of code within the Argus users group and last but not least it avoids re inventing the wheel The routines are catogorized in four different clusters Time processing Argus database access image processing inclusion of fielddata and pixel processing The functioning of each routine is briefly indicated for further details reference is made to help on each function 4 1 Argus time processing Table 4 1 provides an overview of the Argus core routines for the conversion between different time frames of use within the ARE Application epochtime m Returns present epochtime by definition against GMT time epoch2Matlab m Converts epochtime to Matlab s datenum format matlab2Epoch m Converts Matlab s datenum format to epochtime epoch2GMTString m Converts epochtime to a string definition wrt GMT epoch
20. ls2002b 28 Version dd 28 11 02 Appendix D Manual Argus PIXel Toolbox manualArgusTools2002b 29 Version dd 28 11 02
21. ly images which were sampled at a tidal levels between 0 20 m and 0 20 that is the value 0 plus or minus half the range qualify for analysis If the user is interested in storm events only a similar filter can be applied on the basis of wave height information The fielddata water level wave height needed to do so are read from the ARE fielddata archive see Section 3 3 5 Intertidal Beach Mapper IBM IBM is meant for the interactive mapping of shorelines The application allows the user to load an Argus image from the video archive specify a Region of Interest ROD and to determine the location of the shoreline within that ROI For most cameras a default ROI is suggested as well The user interface of IBM is shown in Fig 3 5 manualArgusTools2002b 13 Version dd 28 11 02 erra mele art msp rm Ti aa MA wiere TE TEBEA bonb LI LAST Free trip wa om ieee ss 5 Ape Erg er DT EM gs Fig 3 5 User Interface of the Intertidal Beach Mapper IBM After loading an image from the video archive IBM automatically reads the hydrodynamic conditions from the fielddata archive and the relevant meta information from the argusDB A user specific ROI can be defined by clicking the button Indicate ROI Use the left button of the mouse to specify the corners of the ROI clicking the right mouse button closes the polygon and finishes ROI definition Hitting the Find Waterline button starts the model to determi
22. n separate directories which are typically called delft osu plymouth etc Each database directory contains a file named DB XXXX base 0 mat which provides information on at least one site station image processor camera model and lens model Information on GCP s geometries usedGCP s and eurotide eurowave is stored in separate files for each site The database file of interest is identified on the basis of the siteID which is EGXXX for Egmond DUXXX for Duck etc The database is organized such that is allows for a merging with Oregon State University or any other Argus database e A directory argusTools which contains the various Argus analysis tools discussed in this manual Each application is located in a different subdirectory e A directory CILTools which contains another set of Argus core routines that were originally developed at the Coastal Imaging Lab Oregon State University Many of these routines is used for Argus time processing and the interpretation of image filenames None of these routines however is used for argusDB access see above e A directory fielddata which contains both the fielddata archive files subdirectories tide amp wave as well the Matlab routines that are used to access the information subdirectory code e A directory local which contains user specific core routines Do not remove any of the routines initially provided since they are used at many occasions
23. ne the shoreline This is done by clustering the dry and wet pixels within the ROI in both colour and grayscale space The shoreline is found at the interface of both clusters If the clustering appears to be successful IBM opens a second window showing a 2 dimensional histogram of the two clusters for the criterion colour or grayscale that provides the best distinction Fig 3 6 Besides the identified shoreline is also shown in the Argus image in the IBM main window Discrraradien Dry Wet deers Berei B Habry Dusch Sienai Oath Tg SIE Accept Fig 3 5 IBM Result Window showing the distinction between dry and wet pixels On the basis of a visual inspection of the results the user decides whether to accept the solution provided by the model or to dismiss it If the result largely satisfies apart from a few outliers the user is advised to select Accept After that the outliers can be removed manually with the help of the Remove Dots option Once the result fully satisfies it can be stored by hitting the Store WL button Results are saved in the directory argusTools IBMapper outputDir in anticipation of further post processing The latter may involve the mapping of intertidal beach bathymetry on the basis of set of waterlines sampled over a tidal cycle An example of that is given in Appendix C manualArgusTools2002b 14 Version dd 28 11 02 For a detailed background on the model to identify shore
24. ormation geom cam ip GCPs usedGCPs station site given an image filename shortName like nordzee1 and camera number nordzee 1 Get argusDB station information given station shortName like nordzee1 and epochtime of interest Get all stations of a site given siteID eg NOXXX or site structure Returns usedGCPs from argusDB given sequential number of geometry solution of interest Creates an Argus IP structure for images of non regular size DB2Geometry m Converts geom structure with Walton elements A L stored in separate fields into geom structure with Walton vector m stored in single field geometry2DB m Converts geom structure with Walton vector m stored in single field into geom structure with Walton elements A L in separate fields Table 4 2 Overview of Argus routines for database access The latter 2 routines are needed since the msql database requires the 11 elements A L of the Walton vector to be stored in separate fields while geomtool needs the 11x1 Walton vector m Notice furthermore that geometries returned by DBGetCurrentGeom contain a field m rather than the individual elements A L 4 3 Argus image processing Table 4 3 provides an overview of Argus routines that are typically used for image processing manualArgusTools2002b 17 Version dd 28 11 02 Application argusFilename m Create Argus image filename on the basis of relevant info resin im salon eames ge fora findArgusImage
25. re interpolated If data gaps exist an attempt is made to fill these in with data obtained from the 2 preference source see below following a similar procedure Notice the following aspects The ARE presently accounts for tidal levels wave conditions wave height period and direction and meteorological conditions atmospheric pressure wind Tide wave and meteorological data are located in separate directories under the fieldDataRoot which is defined in argusOpt m The sub directories under fieldDataRoot have to be named Tide Wave and Meteorological Field data are provided as text fiels which obey a fixed format that cannot be changed In the case of tide data this consists of a Nx7 numeric matrix which columnwise represents the years months days hours minutes seconds and the tidal elevation in m Inthe of wave information the data files consists of a Nx 6 n numeric matrix columnwise representing years months days hours minutes seconds and n wave parameters The set of wave parameters can consist of as many parameters as are required A recommended set of wave parameters consists of the rms wave height Hyns in meters the peak period Teak in seconds and the wave angle of incidence degrees with respect to the north positive angles in clockwise direction manualArgusTools2002b 7 Version dd 28 11 02 In the case of meteorological information the data files consists of a N
26. resents the name of the argusDB that has to be initialized For instance argusInit delft Only argusDB s that are specified in the edit section of argusInit m can be initialized Besides argusInit checks whether your ARE license is still valid After running argusInit the argusDB of interest is initialized reported back on screen and the directories which contain the various Argus core routines and Argus analysis tools are added to your Matlab path Notice that e arguslnit can only be run if it is located at your Matlab path Use the option File Set path on Matlab command window menu or the standard Matlab command addpath to add directories to your Matlab path Further details can be found in the Matlab manual e Ifyou are a frequent ARE user it is recommended to include the argusInit call in your startup m file so that your ARE is initialized for every Matlab session Like argusInit m the startup m file has to be located at your Matlab path Matlab Requirements manualArgusTools2002b 22 Version dd 28 11 02 Using the 2002 release of the ARE requires the availability of Matlab 6 1 including the image processing toolbox and the statistics toolbox manualArgusTools2002b 23 Version dd 28 11 02 Appendix B Inputfile Argus Merge Tool AMT Apart from the opportunities offered by the AMT user interface a series of images for merging can also be defined through an external input file Each line of such an input
27. s as cited above Based on the information given in the filename the ARE traces the associated Argus meta information including the most recent geometry solution Notice that only the epochtime number is used for time referencing The time string given in the middle is just meant to facilitate the Argus user who may otherwise have a hard time interpreting epoch numbers 3 2 2 Argus data base of meta information The Argus data base of meta information hereafter referred to as argusDB is a key component of the ARE Without argusDB no quantitative interpretation of image features would be possible Notice the following aspects manualArgusTools2002b 6 Version dd 28 11 02 The argusDB contains all information needed to quantitatively interpret Argus video images This concerns the characteristics of the local site video station image processor cameras applied the available ground control points and geometry solutions etc argusDB can be viewed and edited which may be dangerous with the help of DBOrganizer All meta information is accessible on the basis of the Argus image filename that is epochtime name of the station and camera number The different components of argusDB are all inter connected For instance on the basis of the site identifier all video stations available at that site can found On the basis of the station identifier all cameras of that station can be found On the basis of a camera identifier all geometry
28. s m Find names of Argus images in video archive Several limits rer closest localDay may be applied Read Argus image from video archives imageQuality m Determine image quality Returns good 1 or bad 0 rgb2mono m Convert color image to grayscale ordinates ordinates findUV m Find undistorted image coords UV from field coords xyz findXYZ m Find field coords xyz from undistorted image coords UV Table 4 3 Overview of standard Argus routines for image processing For transformation from image co ordinates to field co ordinates rectification and vice versa bear in mind the following scheme UV coords undistor A undistorted Ra Image Field coords UV coords xyz a UV coords Bee _ undistorted Fig 4 1 Image co ordinates UV gt Field coordinates xyz Fig 4 1 illustrates that image co ordinates UV being distorted need to be undistorted first before being converted into field co ordinates xyz Oppositely fieldcoordinates xyz can be converted into undistorted image co ordinates UV directly For projection on screen however they need to be re distorted 4 4 Accessing external fielddata Table 4 4 provides an overview of routines that are presently used to include fielddata information in the ARE manualArgusTools2002b 18 Version dd 28 11 02 Application DBGetEuroTide m get tidal elevation on the basis of an Argus image filename epochtime of interest or period of
29. solutions available for that camera can be found Etc Using the epochtime of interest as an additional constraint yields a unique geometry solution etc An argusDB is initialized through the function argusInit Example calls are arguslnit delft argusInit plymouth etc See the section above on the initialization of the ARE See Appendix A for an overview of the directory structure of the argusDB s and the contents of the contents of the binary Matlab files found there 3 2 3 External field data Many techniques to derive information from video imagery require the availability of external field data like the tidal level and or the wave conditions The present version of the ARE provides easy fully automated access to external data sources be it that the degree of flexibility is rather limited yet It is important to understand that only the meta information on the fielddata is included in the argusDB The fielddata itself are stored in separate data files ascii format Reading fielddata from file the ARE follows a three step approach 1 2 On the basis of a single time or a longer period of interest the ARE determines from the meta information in the argusDB what fielddata are available Only the files containing fielddata for the time or period of interest are read from file Initially only 1 preference data see below are taken into account To find the data at the time or period of interest the raw field data a
30. te system for Miyazaki Japan Fig 2 1 Argus co ordinate system at Miyazaki Japan The vertical reference level z 0 is generally set to match the mean tidal level or a commonly used often national ordinance level like NAP in the Netherlands 2 2 Argus time conventions Three different time frames are used within the ARE e Epochtime represents the number of seconds since January 1 1970 00 00 00 This is the nine or ten digit number an Argus image filename begins with By nature epochtimes are relative to GMT e Julian days represents the serial number of a day within a year e g Feb 1 32 e Matlab s datenum time frame represents the number of days since January 1 0000 00 00 00 Matlab times are obtained from the standard Matlab function datenum see help on datenum Matlab provides a set of standard functions to convert datenumbers to more accessible formats the most important ones being datevec and datestr See the Matlab help on timefun for further details regarding Matlab time processing Generally it is recommended to do any data processing in epochtime or Matlab s datenum time frame This avoids any troubles with interpolation across different years leap year cycles etc To convert between the different formats the ARE provides a set of standard manualArgusTools2002b 4 Version dd 28 11 02 conversion routines Reference is made to Chapter ffor further details on the routines available
31. tool will project the image area of interest on the ground plane a process called image rectification and the present the result in a separate window e Option Virtual GCP menu GCP virtual gcp This option allows for the definition of virtual GCP s This can be very useful if a site provides clearly visible image objects which are however hard to survey in the field for instance corners of windows or street lights For those cases additional virtual GCP s can be generated on the basis of an existing geometry solution Notice that this should be a high quality solution to prevent the situation that errors strongly accumulate Newly created virtual GCP s are immediately added to the argusDB and stored in an update file It is strongly advised to make only very limited use of this option e Image modification Occasionally poor image contrast or brightness may obscure the proper identification of GCP s For those situations the image menu provides a few options to improve image quality e Option Initial Angles from the selection button in the lower left corner With the help of this option initial values can be defined for each of the 4 angles azimuth roll fov amp tilt that are involved in the iteration process In that way the control over this iteration process is being improved Alternately one can even fix the value of each of the 4 angles Fixed values do not participate to the iteration process hence the n
32. umber of GCP s needed may decrease less unknown variables e Option Pie Pan Pick PPP This option was included to determine the intra pixel location of a GCP on the basis of center of mass considerations It is particularly useful at Duck NC USA where large circular shields were erected to act as a GCP for the Argus station at that site Besides it may be helpful when lights are being used as night time GCP s 3 3 3 Argus Merge Tool AMT AMT provides a user interface to merge images of multiple cameras into panoramic or plan view images Also it can be used to rectify images from a single camera The tool was specifically developed for the merging of large series of images to generate plan view movies or for display at an Argus website manualArgusTools2002b 11 Version dd 28 11 02 ARG Fig 3 3 User Interface of Argus Merge Tool AMT The AMT user interface features 3 different sections The upper left section allows the user to specify the type of merging he is interested in This involves specification of The site of interest egmond nordzeel miyazaki argus02a etc The cameras of interest All or a sub set The input image type snap timex variance or daytimex The output image type colour or grayscale The rectification level So far only the option manual is operational The type of merging Panoramic view or plan view or both On the basis of the site of interest selected here
33. verview of Argus core routines which are indispensable for sophisticated Argus programming The information given here is based on the database format the Argus partners agreed on during the 4 overall Argus workshop organized at Oregon State University in August 2001 Rob Holman Oregon State University USA John Stanley Oregon State University USA and Nathaniel Plant Naval Research Lab MS USA are gratefully acknowledged for their massive contribution to arrive at the operational level of Argus processing where we are today This manual is meant for use with the ARE version 2002b release August 2002 It was written within the framework of the EU sponsored COASTVIEW project under contract number EVK3 CT 2001 00054 Activities for maintenance and further development of the ARE receive co funding from Delft Hydraulics Plymouth University and the Dutch Ministry of Public Works Rijkswaterstaat manualArgusTools2002b 3 Version dd 28 11 02 2 General Argus conventions 2 1 Argus co ordinate system At every Argus site the orientation of the x axis is shorenormal with the positive x axis pointing in seaward direction The y axis is directed perpendicular to the x axis such that the co ordinate system thus obtained is positive in mathematical sense The latter means that the rotation from the x axis towards the y axis indicates the counter clockwise or positive turning direction As an example Fig 2 1 shows the Argus co ordina
34. x 6 n numeric matrix columnwise representing years months days hours minutes seconds and n meteorological parameters The set of meteorological parameters can consist of as many parameters as are required A recommended set of meteorological parameters consists of the atmospheric pressure pam mB wind speed W peea m s and the Wind Direction W degrees with respect to the north positive angles in clockwise direction e To speed up ARE processing ascii data files can be converted into binary Matlab files If available ARE uses the mat files Notice that the mat files use the epochtime time frame rather than the 6 column time definition given in Ascii files Standard routins to convert ascii files into mat files are euroTideAscii2Mat euroWave Ascii2Mat and euroMetAscii2Mat e Since all Argus processing occurs in GMT time field data have to be given against GMT times Correct for a possible time zone offset before adding new fielddata to the ARE e Each data file is represented in argusDB by a Matlab structure which summarizes meta information on that file start time end time columnorder etc On the basis of this meta information the ARE determines the availability of field data for a specific time or period of interest e Thus adding new fielddata to the ARE also means that the Matlab structures of meta information need to be updated otherwise the ARE will not find the new data Being an integral part of the Argus databas
Download Pdf Manuals
Related Search