2BES-WP4.4-D14-Software for the automatization of the
Contents
1. 9 Figure 14 Example of FaceLab Eye lt date gt txt data file 10 Figure 15 Example of the FaceLab World date txt data file seeeeeeeeeeseeeessssss 11 Figure 16 Example of the entree input data file 11 Figure 17 Sample of the enr Dr2 recording parameters file 12 Figure 18 Sample of the var Dr2 output data file 12 Figure 19 The EntreeDr2 intermediate data file 4 13 EIgure 20 FIXATION tme algortlftsssesusdevecvma e does Dvis a Yee e Due sro du PO vae dea cusa e avidus uuu 14 Figure 21 Sample output of the eye tracker s data analysing tool 15 Figure 22 Requestion the recording of playback data 4 16 Figure 23 Eye tracker s replay mode selection ss 17 Figure 24 Sample of the info simu file set for eye tracker s replay mode ssssss 17 Figure 25 Standard display during experimentation ss 18 Figure 26 Simplified view in the eye tracker s replay mode 18 Figure 27 Visual indicators in the simplified view 19 Figure 28 Time offset specification when launching in eye tracker s replay mode 19 LIST OF TABLES Tableau 1 Parameters for the central rear view rendering 14 Document ID 20 Document ID 212. 11091033 var WordPad Seles Fichier Edition Affichage Insertion Format heurecHT indEzsai numlnstruction vp Route vp Vole 10 41 05 17157 12 26900 2007000 O 000 10 41 05 18750 12 2500 2007000 O 000 10 41 05 21675 12 2500 2007000 0 000 10 41 05 25000 12 2000 2007000 O 000 10 41 05 28125 12 2500 2007000 O 000 10 41 05 29687 12 2500 2007000 0 000 10 41 05 32612 12 2500 2007000 O 000 10 41 05 35937 12 2500 2007000 0 000 10 41 05 37500 12 2000 2007000 O 000 10 41 05 40625 12 2500 2007000 O 000 10 41 05 43750 12 2500 2007000 0 000 10 41 05 45312 12 246900 2007000 0 000 10 41 05 45437 12 2500 2007000 O 000 107412905 51552 12 2500 2007000 0 000 10 41 05 53125 12 24600 2007000 O 000 a oo J oc nou whe Po Me H dn e H r2 Li OO OO COQ OO OO eo oo oo OO OO OO H Ha Appuyez sur Fi pour obtenir de l aide Figure 18 Sample of the var Dr2 output data file Document ID 12 2 BE SAFE D14 Software for the automatization of the analysis of eye tracker s data 2 3 Input data merging and analysing 2 3 1 Pre processing One problem with this arrangement is that the eye tracker has a fixed acquisition rate of 60 Hz while the simulator has an average frequency of 60 Hz that may fluctuate during the experiment It was therefore necessary to combine the data from both sources intelligently and automatically according to a common time The merging process assumes that both the
3. panneau 0 1 80 30000 means that the area of interest will be linked to the object 0 located on road number 1 at kilometric point KP 3 0000 and located at a lateral position of 8 meters The face category allows creating a zone of fixed interest whose location is specified by the 4 vertices of the face Syntax face name list of the co ordinates of the 4 vertices face building 1 674 932050 34 8886820 13 5751250 674 024390 34 9777690 3 2048E 01 651 215930 34 6028970 2 5106E 01 654 450030 34 6420930 13 4708410 means that the area of interest consists in a polygon defined by the co ordinates which are pr cised Document ID 4 2 BE SAFE D14 Software for the automatization of the analysis of eye tracker s data There is no obligation to design a rectangle The interior angles do not need to be straight An additional module has been added to the SIM application which provides an interface to select and specify the various parameters involved To access them you must 1 Load and display your 3D base by choosing the right track in SIM 2 Not be connected with DR2 which mean off line use of SIM 3 Select helicopter view by pressing F9 Figure 2 is the dialog box which allows to precise the areas of interest Actions Objets Marquage Echantillonnage f Creation Bee t Destruction Nouvelle Route f D placement Biri D placement zone d interet Skatique Whe m fs m Panneau Commenta
4. q CE AFE 7 2 WHEELER BEHAVIOUR AND SAFETY ERR SEVENTH FRAMEWORK PROGRAMME 2 BE SAFE 2 WHEELER BEHAVIOUR AND SAFETY Grant Agreement number 218703 Funding Scheme Collaborative project small scale Start date of the contract 01 01 2009 Project website address www 2besafe eu Software for the automatization of the analysis of eye tracker s data while used with the driving simulator Confidentiality level restricted Deliverable D14 WP Activity 4 4 Status Final version Date 14 06 2010 Author INRETS Project Coordinator Scientific Coordinator St phane LAPORTE St phane ESPI Europe Recherche Transport INRETS c o LCPC 2 rue de la Butte Verte 58 Bd Lef bvre F 93166 Noisy le Grand Cedex F 75732 Paris Cedex 15 Stephane laporte ert sas fr espie inrets fr Phone 33 0 1 45 92 55 23 Phone 33 0 1 40 43 65 54 Authors Company INRETS Name F Vienne S Espi Amendments Date of Issue Description 07 05 2010 First draft by INRETS 14 06 2010 Modifications following the peer review Applicable Documents Description comments Acknowledgements Description comments Document ID file name 2 23 Table of contents 1 Software for the automatization of the analysis of eye tracker s data sseesseuusss 3 1 1 Identify static and dynamics areas of interest 3 1 1 1 Areas of interest VEMICICS wisiierentshecnawiscaesaachcvtesnnt
5. 0 05 785 090 10 42 49 695 00 00 13 649 090 10 42 51 527 00 00 16 856 016 10 42 52 062 00 00 17 392 015 10 42 52 195 00 00 17 512 021 10 42 53 878 00 00 20 725 035 10 42 53 945 00 00 20 805 016 10 42 57 992 00 00 27 053 117 10 42 58 894 00 00 20 901 139 10 43 04 007 00 00 37 190 Oo 10 45 07 890 00 00 43 743 055 10 23 07 957 00 00 44 020 2275 10 43 15 585 00 00 56 575 019 10 43 15 757 00 00 56 740 016 10 45 16 524 00 00 57 374 034 10 43 18 538 00 01 01 097 013 AA A aa mM md ra n rr Appuyez sur F1 pour obtenir de l aide J208o000o0030rnudocooco dcocz3uod 7 D mn S mn D m D m S m S m SO m G m G m G m D m D a E m DA a D m A a n Figure 21 Sample output of the eye tracker s data analysing tool Document ID 15 2 BE SAFE D14 Software for the automatization of the analysis of eye tracker s data 2 4 Operating mode 2 4 1 DR2 traffic simulation In order to create DR2 playback files to be used in the eye tracker s replay mode you have to request the creation of the following three files when launching your experiment e res mdv bin recording the subject s actions on the controls pedals steering wheel gearbox so you can replay them later e res monde bin traffic information e res env bin environnemental information time These are binary files To generate these files you have to tick Enregistrement as shown below mode de fonctionnement calibrage des codeu
6. 100261 1 100262 i 100263 ed HEAD OBJ INDEX HEAD OBJ NAME CYC SCR INTSECT nothing nothing nothing nothing nothing nothing nothing nothing nothing nothing nothing nothing nothing nothing nothing nothing nothing nothing nothing nothing nothing nothing not hi net 100264 100265 100266 100267 L002 66 1 L002 69 a 100270 100271 100272 1 100273 1 2 LR oH H H H H HR H Ho LR HR RoOR H R oH Ho Ro HR H BH 100774 1 TIT Appuyez sur F1 pour obtenir de l aide Figure 15 Example of the FaceLab World date txt data file The software thus creates an input file for the eye tracker s values including the aforementioned 6 data Fri Sep 11 2009 10 42 43 entree WordPad ME Fichier Edition Affichage Insertion Format GMT S GMT MS 1252665761 563 1252665761 579 GAZE QUAL R GAZE QUAL L GSI SCREEN X 2 0 0797376 0 0558399 GSI SCREEN Y 0 472495 E 0 399988 k Appuyez sur Fi pour obtenir de l aide Document ID 1252665761 1252665761 1252665761 1252665761 1252665761 1252665761 1252665761 1252665761 1252665761 1252665761 1252665761 1252665761 1252665761 1252665761 aes ccmnmnc a 596 613 629 646 663 679 696 713 Tad T46 TOS TTS 796 815 1h PS D3 Dh D3 P3 D3 D3 Po P3 D3 D3 D3 bo bo bo P3 MM ES MW P3 D3 bh Es oh D bh Bo bo M O O O O 0 0 O O O O O O O O 00183827 059
7. 62 00795615 0 016182 00 0460203 0 0345955 n naini5ss 0 344524 0 44636 0 407678 0 289435 0 31503 0 031555 O 03 71423 0 033969 0 033042 5 O 0417664 0 031134 O 03 60255 0 031412 O 03 64647 0 033951 O 0260118 0 0292957 0 0304495 0 031939 0 0331935 0 033245 0 050397 n n aama5 Figure 14 Example of FaceLab Eye date txt data file 2 1 2 The FaceLab Timing date txt data file This file contains 2 values GMT S and GMT MS GMT S is the GMT date in seconds of each acquisition GMT MS corresponds to the part in milliseconds to add to the previous value A utility allows to rewrite the data in HH MM SS sss notation to compare with the time from the DR2 recordings Document ID 10 2 BE SAFE D14 Software for the automatization of the analysis of eye tracker s data 2 1 3 The FaceLab World date txt data file This file contains 2 values GSI WORLD X and GSI WORLD Y which give the FaceLab X and Y co ordinates on the virtual screen see Figure 15 The virtual screen is determined directly by the FaceLab system Its dimensions have been integrated into our software in order to achieve a relevant normalisation and therefore a good positioning on one of our three screens Bei B World_Fri_Sep_11_2009_10_42_43 txt WordPad Fichier Edition Affichage Insertion Format FRAME NUM 100252 1 100253 i 100254 i 100255 i 100256 i 100257 i 100258 i 100259 i 100260 i1
8. 7621 0838035 0642931 054471 0544751 0845317 0545759 0653259 05460777 062966 0620194 0507551 0795514 0 225001 0 0491755 0240207 024297 02435647 0204261 0161584 018192 0199902 0154941 00671503 00285368 0115596 0191691 amam m o S m S m B m S ma G m G oo 8 Figure 16 Example of the entree input data file 11 2 BE SAFE D14 Software for the automatization of the analysis of eye tracker s data 2 2 DR2 output data file The DR2 software allows to precise the parameters which have to be recorded Figure 17 during an experiment The recording period is indicative and may slightly fluctuate during the recording due to for example a slowdown in the visual rendering process BP 2besafe enr Bloc notes Fichier Edition Format Affichage F Les colonnes suivantes contiennent les noms des donn es enrega gt Vit posXx posy F P riode E 00 P riodesauwegar de simu 2 HeureaGMT pas Scenario 2 MLumrnmnst IndEessai Vitesse cab rndics Vitesse Vitesse Vitesse vitesse vitesse vitesse vitesse BREE EE Ein Figure 17 Sample of the enr Dr2 recording parameters file The output file Figure 18 contains the requested information In the example the step number of the simulation acquisition time and many other data specific to the experiment among them the position of the subject vehicle s on the road network and of 25 traffic vehicles he interacts with
9. KP specification 1 1 3 Areas of interest 3D polygons In the case of a face one has to press the space bar when the mouse cursor is in the right positions The application determines the first polygons encountered in the virtual world automatically When the fourth and last vertex 1s specified the following dialog appears on the screen Figure 8 Mom de la face 15 carac max Figure 8 Face name specification This allows the user to remember more easily what this face represents One will then be ask to confirm the choice of name here tt Figure 9 Document ID 7 2 BE SAFE D14 Software for the automatization of the analysis of eye tracker s data l ectBase ChargeFichiet 2 voulez vous conserver la Face tt Figure 9 Confirmation for a face area of interest If one click on Oui the face specification is stored in the file otherwise not 1 2 Displaying the areas of interest Here is an example of an initial view of the driving area with two objects e A dynamic car number 0 e A static a speed limit sign position route 17 lane position 40 dm KP 14900 dm Figure 10 Initial view taken during the experiment We thus want to create two zones of interest one for each object This means creating a file containing all the required information Figure 11 shows such an ASCII file Fichier Edition Affichage Insertion Format vhs panneau 17 il 14900 Appuyez sur F1 pour o
10. btenir de l aide Figure 11 Example of Zonelnteret parameter file Document ID 8 2 BE SAFE D14 Software for the automatization of the analysis of eye tracker s data When one executes the application in eye tracker s replay mode one gets the following display Figure 12 HeureGMT VAR 10 42 54 76562 HeureGMT OCULO 10 42 54 777 Figure 12 Eye tracker s replay mode Note that both objects are now encompassed by rectangular parallelepipeds that hug the shapes of objects The notion of distance between the faces nearest to the object is very important In fact the calculation of fixation durations in each zone is based on the intersection of the measured gaze with the parallelepiped An error may be induced if the object is too broadly embraced A considerable amount of work has been done to minimise these errors Note the entire visual base is completely white no textures except for the areas of interest 2 Input files of the application The application which analyse visual activity requires three input files The first comes from the merging of the eye tracking files 1 the second from the data generated by DR2 2 and the third is the previously defined parameters file describing the areas of interest Figure 13 ge e pum 1 i nn y ER Figure 13 Schematic representation of the application s inputs and outputs Document ID 9 2 BE SAFE D14 Software for the automatization o
11. f the analysis of eye tracker s data 2 1 Eye tracker s output data files The FaceLab eye tracker generates four files The first one lt date gt entree contains the information of interest from the three other txt files The name of each of the files generated by the eye tracker precise the date of the experiment The lt date gt string being composed of Weekday month day year hour minute second corresponding to the start of the eye tracking acquisition e g Timing Tue Sep 8 2009 14 26 36 txt The following sections details each data file generated by the eye tracker 2 1 1 The FaceLab Eye date txt data file This file contains 2 values GAZE QUAL R and GAZE QUAL L see Figure 14 These correspond to the pupil detection quality A value of 2 is required to use the position values This 1s a filter in the treatment of eye tracking data Z Eye Fri Sep 11 7009 10 42 43 txt WordPad Fichier Edition Affichage Insertion Format FRAME HUM EIGHT EYE CLOSE LEFT EYE CLOSE RIGHT CLOS CONF Appuyez sur Fl pour obtenir de l aide 1uuzs5z 100253 100254 100255 100256 100257 100255 100259 100 60 100 61 1002 62 1002 63 100264 100265 100266 100267 L002 66 100269 100270 100271 100272 100273 1002774 0 161591 0 115645 0 162455 0 116146 0 116313 0 0514105 E OO 0 104557 0 152077 O 07 78462 0 026913 Dn n0565473 D 0714821 D 04374 7 E 0 0 003199
12. ile generated by DR2 e HeureGMT oculo the GMT time from the eyetracker e A cursor indicating the gaze Note the difference between the GMT times should never exceed 16ms Document ID 18 2 BE SAFE D14 Software for the automatization of the analysis of eye tracker s data HeureGMT VAR 10 42 54 289 e HeureGMT OCULO 10 42 54 77 e Figure 27 Visual indicators in the simplified view In order to configure the device in the best possible way there is a possibility of adding an offset time between the timer of the DR2 and the eye tracking PCs This allows compensating for a lack of synchronisation between the two clocks at the launch of the application in the eye tracker s replay mode Offset s entre PC Oculo et PC Dr gt 0 si PC Oculo en avance x 22 Cancel fe Figure 28 Time offset specification when launching in eye tracker s replay mode 3 Conclusions The aim of the 4 4 activities 3 was the design of a software tool for the automatization of eye tracker s data while used with the driving simulator This activity has been conducted successfully and the software was delivered to researchers of WP5 2 The method used for the automatization is a post treatment process which allows researchers to analyse their recorded data and to easily change the interest areas The areas may consist in either static or dynamic objects The areas can be easily described using symbolic description or geo
13. ire t Orientation Face Afficher Trac Figure 2 Areas of interest selection dialog box This dialog box is used for multiple purposes only the zone d interet part bottom right is used for identifying the areas of interest 1 1 1 Areas of interest vehicles If Vhs vehicles is checked then appears on the screen a new dialog box Figure 3 Document ID 5 2 BE SAFE D14 Software for the automatization of the analysis of eye tracker s data zone d interet numero du vehicule Figure 3 Vehicle ID number specification One must specify the DR2 identification number of the vehicle concerned 1 1 2 Areas of interest objects If panneau objects is checked the following dialogs will appear Figure 4 zone d interet numero id du panneau Figure 4 panneau ID type specification One must specify the position of the SIM ID type for the object Then one has to specify a position First one specifies the road s identification number DR2 reference Figure 5 zone dinteret numero de route du panneau Figure 5 Road ID number specification Document ID 6 2 BE SAFE D14 Software for the automatization of the analysis of eye tracker s data As well as its lateral position 1n dm Figure 6 zone dinteret voie du panneau Figure 6 Lateral position specification And at last it s KP in dm Figure 7 zone d interet pk du panneau Figure 7
14. metric faces If the calibration of the eye tracker has been properly conducted the software can be used in a reliable and efficient way The produced software has been designed for the INRETS SIM class simulators and for French users which are the users within 2BeSafe but the method can be fairly easily adapted to other driving simulators as well as the man machine interface language Document ID 19 2 BE SAFE D14 Software for the automatization of the analysis of eye tracker s data LIST OF FIGURE Figure 1 Symbolic description of areas of interest vehicles and signs sssseeeusssss 4 Figure 2 Areas of interest selection dialog box Rs 5 Figure 3 Vehicle ID number specification sis 6 Figure 4 panneau ID type specification iii 6 Figure 5 Road ID number specification siennes 6 Figure 6 Eateral posillorspeclliCallOlisssse cubes cosa ttp UO Eia aa t eu ue eau a qu bue dca aM M mE RVERU VERA 7 Figure 7 s KP SD CINCAION ins nn Malada ee uetus Mult npo ac vn Nd bud AR cn ts 7 Figure 8 Face narme sDeclHCallOliasscuc icr AR Te d cda pod eet bea Dil pr ic eu cv pudo Reus 7 Figure 9 Confirmation for a face area of interest 8 Figure 10 Initial view taken during the experiment ss 8 Figure 11 Example of Zonelnteret parameter file 8 Figure 12 Eye tracker s replay mode in 9 Figure 13 Schematic representation of the application s inputs and outputs
15. of 0 05 Red the second 0 1 etc Document ID 13 2 BE SAFE D14 Software for the automatization of the analysis of eye tracker s data The algorithm can be divided into five steps Step 1 Retrieval of the X and Y position co ordinates of the FaceLab virtual window This is done using information from the file date entree Step 2 Normalisation of these values to place them in the SIM environment Identification of which of the three views left centre right contains the gaze Step 3 Projection of this 2D coordinates display into the virtual world using ray tracing to find the first polygon encountered in the 3D world Step 4 Testing the colour of the intersected polygon for its Red component The value is compared with that of different areas until a match 1s found Step 5 Two possible cases The gaze was not within this AI at the previous time step A new fixation period is thus created The gaze was within this AI at the previous time step The duration of the current fixation period is thus updated Duration of the current fixation step t 1 Duration of the current fixation step t time between the two steps t and t 1 Figure 20 describe the fixation algorithm In this case one line inside the tempsOrigine t 1 output file dureeFixation time between t 1 t 2 time between t 2 t 3 D E E E i t t 1 2 doi t 4 O O O tempsOrigine t 1 tempsOrigine t 3 dureeFixation time be
16. r s and data from the driving simulator are recorded 2 Determining the elements or areas of interest for the visual activity Three kinds of elements can be defined vehicles road signs and areas Vehicles and road signs are identified symbolically while areas are defined using the coordinates of 3D rectangle faces The defined areas can be double checked visually using the driving simulator visual loop 3 Analysing the data using a replay mode which allows correlating the data acquired by the eye tracker s the data acquired by the driving simulator and the data defined for the study of the visual activity It has to be stressed that the step 1 can be achieved before the step 2 but has of course to be achieved before the step 3 It has also to be noticed that the design of the tools in such a way allows the processing of the data acquired in previous experiments The following sections explain the process and may be used as a user manual As the software is primarily intended to be used by French users the language used 1s the French 1 1 Identify static and dynamics areas of interest The software allows selecting up to 20 areas of interest for the experiment This characteristic has been defined jointly with the researchers of WP5 2 it can be easily increased up to 255 Three categories of areas of interest can be specified e vhs vehicles e panneau sign or object e face polygon Document ID 3 2 BE SAFE D14 Sof
17. rs variateur analogique W prise en compte z ro volant gestion si ge vibrant dbg Cabine Chois du mode f mode 0 REY calcul par PC envoy par PC mode 1 courbe Assistance seule t l charg e mode 2 courbe Rappel seule t l charg e f made 3 courbes Rappel et Assistance t l charg es Chom du mode de pilotage fichier surcharge cabine manette rejeu avec Alarmes m f cabine m Z reyeu UDF Annuler Figure 22 Requestion the recording of playback data When you want to obtain your output file with the fixation times on the different areas of interest you must specify that you are in replay mode Document ID 16 2 BE SAFE D14 Software for the automatization of the analysis of eye tracker s data mode de fonctionnement calibrage des codeurs variateur analogique Iw prise en compte z ro volant gestion si ge vibrant dbg Cabine Chois du mode mode 0 REY calcul par PC envove par PC mode 1 courbe Assistance seule t l charg e mode z courbe Rappel seule t l charg e made 3 courbes Rappel et Assistance t l charg es Chom du mode de pilotage fichier surcharge cabine manette enregistrement m Z ejen avec lames f rejeu UDP Annuler Figure 23 Eye tracker s replay mode selection 2 4 2 SIM driving simulator visual loop To display the eye tracking mode in Sim all white with the ZI in shades of red you have
18. sivainatsunesseds cassapbssbesnudsinsiaablunciadiccnesaskexns 5 1 1 2 Areas of interest objects iii 6 1 1 3 Areas of interest 3D polygons ss 7 1 2 Displaying the areas of interest nnns 8 2 WADUT MES OF INS ADDIICATIOMN o 9 2 1 Eye tracker s output data files 10 2 1 1 The FaceLab Eye date txt data file 4 10 2 1 2 The FaceLab Timing date txt data file 10 ANS The FaceLab World date txt data file 11 2 2 DRZ outputaata Te RR 12 2 3 Input data merging and ANALYSING cccccseecceccaeeseeeeeceeeeeceeeceeeeeeeesseeseeeeessaaeeeeeeseaneeeeeeeas 13 2 3 1 Feno eS Ieee E E E E E 13 2 9 2 Data ES LA SUN ES CR PR ET E Ea EE a E R LT 13 2 3 3 Output file of the analysis software 15 2 4 Operating Mode PM 16 2 4 1 Biaxicuier necu 16 2 4 2 SIM driving simulator visual loop 17 2 5 VISDAMIMO dicus t NETTE UT Umm 18 SEE dro geile gm m 19 MGNS ICI T en 20 BIN Ncc 20 Document ID 1 2 BE SAFE D14 Software for the automatization of the analysis of eye tracker s data Executive summary This report documents activities 4 4 Car simulator sub activities 3 which objective is the design of a software for
19. step number of the simulation and the acquisition time has been recorded It also assumes that each step of the simulation is recorded PeriodeSauvegarde has to be set at 1 A file EntreeDr2 Figure 19 is created based on the DR2 output file A column corresponding to the number of seconds and milliseconds since the start of the day is added for the purpose of comparison with the eye tracking data E Fri Sep 11 2009 10 42 43 entreeDr2 WordPad ML Ej Fichier Edition Affichage Insertion Format HeureGNT 3s GMT 5 journ e 10 41 05 171867 30465 170 10 41 05 18750 30465 187 10 41 05 21875 30465 2165 10 41 05 25000 30465 2450 10 41 05 28125 30465 200 10 41 05 29687 36465 297 10 41 05 3231Z 38265 3z28 10 41 05 35937 30465 3568 10 41 05 37500 30465 3775 10 41 05 40625 38265 4085 10 10 21 05 23750 38265 2337 Appuvez sur F1 pour obtenir de l aide 1 2 3 E 5 B 7 d g Figure 19 The EntreeDr2 intermediate data file 2 3 2 Data analysis The analysis uses the SIM visual loop but in a simplified mode The entire environment is white textures are not displayed except for the areas of interest The area of interest which have been specified by the user see 1 1 are displayed using bounding boxes with a specific uniform colour which is created automatically The colour code is R XX G 0 B 0 where the R value is used to differentiate the AI Areas of Interest The first area has a value
20. the automatization of the analysis of eye tracker s data while used with the driving simulator This tool will be used in WP5 2 for conducting experiments on car simulator about rider s conspicuity The method used for the automatization is a post treatment process which allows researchers to analyse their recorded data and to easily change the interest areas The areas may consist in either static or dynamic objects The areas can be easily described using symbolic description or geometric faces If the calibration of the eye tracker has been properly conducted the software can be used in a reliable and efficient way The produced software has been designed for the INRETS SIM class simulators and for French users which are the users within 2BeSafe but the method can be fairly easily adapted to other driving simulators as well as the man machine interface language As a final result the software has been designed implemented and delivered it fit the user requirements Document ID 2 2 BE SAFE D14 Software for the automatization of the analysis of eye tracker s data 1 Software for the automatization of the analysis of eye tracker s data The aim of this activity was the design of tools which helps the final user in setting and operating the results of their experiments using eye tracker s on simulators Three integrated tools have been designed 1 An off line one allowing to precise elements in the visual scene which have
21. to modify the value of the enregistrementOculo option in the info simu file as shown in figure 23 info simu WordPad Fichier Edition Affichage Insertion Format afxMeteo stfxMeteoIntensite voayeurPur noFocus 1 delai Oo messageFichier E filtre 0 000 0 000 0 000 0 000 acLivationc E connectioniuto eclairagePublic mireOculometrre enregistrementoculo Num ro de version niveau de notification Performer 0 7 messages d boque programme 0 1 phaseFixe 0 11 1 limite par freqimagqe 1 la plus rapide possible freqimage 1 tailleFenetre x vy l H mode de duplication des mobiles O gt duplication par clonage flatter EH H HHHH HH Appuyez sur F1 pour obtenir de l aide Figure 24 Sample of the info simu file set for eye tracker s replay mode Document ID 17 2 BE SAFE D14 Software for the automatization of the analysis of eye tracker s data If this option is set to 0 you will obtain a standard non replay eye tracking display as shown below Figure 25 Standard display during experimentation If the option is set to 2 one obtains the eye tracker s replay mode HeureGMT VAR 10 42 54 76562 HeureGMT OCULO 10 42 54 777 Figure 26 Simplified view in the eye tracker s replay mode 2 5 Visual information The software provides several visual indicators directly on the Sim interface These are e HeureGMT var the GMT time obtained from the var f
22. to be considered for the visual activity 2 Anon line one aiming at synchronizing the data acquired by the simulator and them acquired by the eye tracker s 3 An off line one which allows calculating the duration of the fixation point object in space at which sight is aimed for each considered elements It has to be considered that such tools are to a large extent simulator and eye tracker s dependant as the 3D database format the 3D rendering engine and the eye tracker s manufacturer are often different from simulator to simulator Thus the tools designed in this activity are dedicated to the simulator used in WP5 2 which is a INRETS SIM class simulator and to the FaceLab eye tracker s The INRETS SIM simulator architecture is an open and modular architecture which 1s used in several research labs in France and also in some foreign research labs Compare to existing systems for example the Oktal one the tools have been designed in such a way that they simplify the configuration and the analysis by non expert users The first step consisting in defining the elements considered for the visual activity do not require skills in info graphics as the user can precise areas in the 3D database without using a 3D modeller The process which assumes that the protocol of the experiment has been properly defined consists in 1 Achieving the experiment using the simulator and the FaceLab eye tracker s in this step data from eye tracke
23. tware for the automatization of the analysis of eye tracker s data These data have to be specified in an ASCII file named Zonelnteret txt Figure 1 1s an example of such a file Zonelnteret txt WordPad Fichier Edition Affichage Insertion Format whe O panneau Appuyez sur F1 pour obtenir de l aide Figure 1 Symbolic description of areas of interest vehicles and signs The vhs vehicles category creates a dynamic area moving with the vehicle of which you specify the identification number within the simulation The identification number consists in the DR2 ID of the vehicle DR2 being the software in charge of the traffic simulation within the INRETS SIM simulator architecture Syntax vhs DR2 ID of the vehicle vhs 0 means that we determine the time of fixation on vehicle 0 when it is displayed The panneau sign or object category allows to create an area of interest related to an object from the static background road sign tree house you specify the identification number within the simulation The identification number consists in the SIM type of the object as the objects of this category are not always managed by the traffic simulation In order to avoid confusion the location of the object is added using the curvilinear position of the object along a specific road Syntax panneau SIM ID of the object road number lateral position of the object dm kilometric position of the object dm
24. tween t 1 t 2 dureeFixation time between t 3 t 4 In this case two lines inside the output file Figure 20 Fixation time algorithm Document ID 14 2 BE SAFE D14 Software for the automatization of the analysis of eye tracker s data 2 3 3 Output file of the analysis software At the end of the eye tracker data analysis process the application has generated an output file containing all the areas of interest set during the experiment This file Figure 21 specifies e tempsOrigineOculoGMT to refer to the eye tracker input files This gives the possibility to carry out offline cross comparisons e tempsOrigine time since the beginning of the experiment Value obtained from the simulator data e noZone the number of the zone of interest Remember this number depends on the line number in the file ZonelInteret txt the first line being number 0 e pas the step number in DR2 s iterative loop Also provides a reference to synchronise with respect to the var file Of interest mainly in relation to scenario data With this we may associate the simulation time and the experimentation timing the position in scenario e dureeFixation time during which the eye fixed the same IA without leaving this area E Fri 5ep 11 2009 10 42 43 sortie WordPad Fichier Edition Affichage Insertion Format tempsoridgineOculocHT tempsoricdgine nozone dureeFixation 10 42 41 945 00 00 03 197 036 10 42 44 527 00 0
Download Pdf Manuals
Related Search