Selection of the basic parameters of the lens for the optic-electronic target recognition system

Transcription

1 Proceedigs of the 5th WSEAS It. Cof. o COMPUTATIONAL INTELLIGENCE, MAN-MACHINE SYSTEMS AND CYBERNETICS, Veice, Italy, November 0-, Selectio of the basic parameters of the les for the optic-electroic target system MARTIN MACKO - TEODOR BALÁŽ Departmet of Weapos ad Ammuitio Uiversity of Defece Bro, Kouicova 65, 6100 CZECH REPUBLIC marti.macko@uob.cz teodor.balaz@uob.cz Abstract: This article deals with determiatio of les parameters of the optic-electroic system for detectio, ad of movig military targets i depedecy o their distace. There are basic assumptios for determiatio of the most importat parameters of the leses suitable for systems of automatic target acquisitio described i the article. Key-Words: Les, camera, image, military target, detectio,, 1 Itroductio Observatio ad automatic processig of the scee at the military applicatios differs from the idustrial applicatios especially at the distace of observed movig objects military targets. The optical part of the itelliget system for observatio of the combat scee must allow trackig of movig targets up to distace of 3000m. This big rage of distaces of the target creates very high demads o quality of the optical system, the optic-electroic system, ad also the capability of the algorithm to recogize ad idetify the military target. The "surveillace path" of the optic-electroic system for object cosists of surveillace camera with les, electroic uit ad display uit. The system surveillace path computer-software algorithm allows detectio,, ad of the target. By meas of this system is a ragefider guided o the target to localize him. Problem - Detectio, ad of the target The most importat parameter for evaluatio of quality of optic-electroic system for target, from user poit of view, is his rage for idividual types of targets. The rage of the surveillace path is the maximum distace of the target at which ca be detected, recogised ad idetified. Detectio, ad are three levels of recoaissace task, see Table 1. criterio detectio Table 1 Resolvig criteria descriptio of image of the target patch, target will bee see target recogised, distiguished class (tak, armour fightig vehicle, etc.) target recogised, specified type (tak T-7, versio M1) 3 Solutio of the problem The rage of optic-electroic system for target ad quality of image of the target are always defied o the basis of the followig coditios ad parameters: emissive ad reflective characteristics of the target, his effective surface area ad orietatio; distace of the target ad his cotrast towards backgroud; trasfer characteristics of atmosphere; focal legth of sca les ad its focal ratio; quality of optical imagig; geometric characteristics of detector; sesitivity of detector; miimal illumiatio, exposure time. O basis of may experimets, e.g. [4], [7] was cocluded that the most suitable parameter for evaluatio of accomplishmet of give visual

2 Proceedigs of the 5th WSEAS It. Cof. o COMPUTATIONAL INTELLIGENCE, MAN-MACHINE SYSTEMS AND CYBERNETICS, Veice, Italy, November 0-, recoaissace task, i. e. detectio, or of the target, is the bar-chart resolutio test. The measures ad criteria of performace of resolutio were origially developed by Johso at Today is this method still improved ad elaborated. Spatial frequecy (period), dimesios, mea magitude of lumiace ad cotrast of bar-chart resolutio test depeds o solved recoaissace task, type of tracked target, ad coditios of observatio. Dimesios are the same as critical dimesios of the target (the smallest dimesio of the target). Mea magitude of lumiace L depeds o both mea magitude of lumiace of target L S, C ad backgroud L S, P :, C +, P, T =. (1) Cotrast of bar-chart resolutio test is give by:, C, P LT C Δ T = =. () L L Accordig to [4] holds true for determiatio of probability of detectio, ad i depedecy o umber of periods at equivalet bar-chart resolutio test across target of critical dimesio followig relatios: E P 50 =, E (3) E = , (4) where is a umber of periods at equivalet barchart resolutio test ad 50 is umber of periods at equivalet bar-chart resolutio test for 50% probability of detectio, ad. Idividual probabilities are show at Figure. 50 Fig. 1 Equivalet bar-chart resolutio test for determiatio of 50% probability of detectio, ad Table 50% probability of solvig recoaissace task umber of periods of equivalet bar-chart resolutio test o target critical dimesio of target detectio tak AFV car perso howitzer

3 Proceedigs of the 5th WSEAS It. Cof. o COMPUTATIONAL INTELLIGENCE, MAN-MACHINE SYSTEMS AND CYBERNETICS, Veice, Italy, November 0-, detectio umber of periods Fig. Idividual probabilities of detectio, ad of the target i depedecy o umber of periods of equivalet bar-chart resolutio test Table 3 50% probability of solvig recoaissace task for target of height of 3m i depedecy o focal legth of scaig les f [mm] ω v [ ] ω s [ ] detectio O the basis of calculatio of the 50% probability solvig of recoaissace task for target of height of 3m i depedecy o variable focus legth of scaig les were obtaied the rages for detectio, ad for seve differet leses, see Table 3. From Table 3 follows that for real observatio ca be used the les with focal legth f = 50 mm. The surveillace path ca secure, i case of size oe pixel of the surveillace CCD camera ρ = 8 μm, detectio of above defied target at distace of 9375m, at distace of 343m ad at distace of 1171m.

4 Proceedigs of the 5th WSEAS It. Cof. o COMPUTATIONAL INTELLIGENCE, MAN-MACHINE SYSTEMS AND CYBERNETICS, Veice, Italy, November 0-, Fig. 3 Example of modellig of, ad detectio of predefied target at differet distaces at imagig of surveillace path The Figure 3 shows images, as they appear at surveillace path, from modellig of, ad detectio of the defied target at differet distaces. The distaces preseted at Table 4 are valid for idividual levels of recoaissace task oly uder assumptio ideal visibility ad itroduce "geometrical" criteria of maximum rage. Cotrast betwee target ad backgroud is c = Durig determiatio of basic parameters of the les it is ecessary to set the miimal focal ratio c. The selectio of parameters of leses is govered maily by required quality of image of the target ad must respect followig physical limits. I case whe oe pixel of the detector is a square of side d, the focal ratio c a of les which Airy disc just fits ito this pixel holds true followig relatio: da ca =. (5).44 λ Takig ito accout size of the pixel 8 μm x 8 μm, wave legth λ = 600m, the miimal focal ratio of the les will be about 5.6. Aother importat characteristic of the les is its optical trasfer fuctio of cotrast. I case of ocoheret radiatio, all cases of scaig of military armamet ad vehicles, is the trasfer fuctio give by followig relatio: D( ) = ( arccos s s 1 s s ), (6) π where s= λ c R ad R is real spatial frequecy at image plae. This characteristic is very importat for determiatio of quality of imagig o the detector. For example whe the tak of height of.5m is at the distace of 3000m from the camera its spatial frequecy is 100lies/rad ad is imaged by les with foal ratio f = 5.6 for wave legth λ = 600m, cotrast k = 0.88, see poit d at Figure 4. I case of task whe it is required to resolve 7 periods is the spatial frequecy 8400lies/rad ad they are imaged with cotrast c = 0.7, see poit r at Figure 4. Whe cotrast of the tak with regard to the backgroud is c = 0. 6 ad cotrast of Johso criteria is c = 0.7 the durig detectio the tak will appear with cotrast c = ad cotrast betwee periods of bar-chart resolutio test will be c = which is o boudary of.

5 Proceedigs of the 5th WSEAS It. Cof. o COMPUTATIONAL INTELLIGENCE, MAN-MACHINE SYSTEMS AND CYBERNETICS, Veice, Italy, November 0-, Fig. 4 Optical trasfer fuctio of les cotrast c = 5, 6 for wave legth λ = 600m 4 Coclusio There is show way of determiatio of basic parameters of les. The determiatio of these parameters belogs to the first step at sythesis of optic-electroic system for target trackig. These parameters allow, at least approximately, determie quality of image of the target at detector plae ad cosequetly determie probability of detectio, ad. Refereces: [1] V. Molebyj, Optiko-lokacioye sistemy, Moskva, Mašiostrojeie, [] M. Bass, Hadbook of optics, Vol. 1,, McGraw-Hill, New York, [3] J. Jakušekov, Teoria i rasčot optikoelektroych priborov, Mašiostrojeie, Moskva, [4] G. Holst, Electro-optical imagig system performace, d Editio, JCD Publishig, Witer Park, 000. [5] P. Fischer ad J. Bucholcer, Tak fire cotrol systems, Vol., Military academy, Bro, 00. [6] H. Haferkor, Bewertug optischer Systeme, DVW, Berli, [7] R.G. Driggers, P. Cox ad M. Kelley, Natioal iterpretatio ratig system ad the probabilities of detectio, ad, Optical Egieerig, Vol.36, No.7, 1997, pp [8] Schlesiger M.I., Hlaváč V.: Te lectures o statistical ad sytactic patter, Kluwer Academic Publishers, Dordrecht, The Netherlads, 00, 51 p. [9] Vítek, R.-Macko, M.-Baláž, T.: Techologie sběru fotogrammetrických dat kamerovým systémem, Výzkumý záměr FVT , Uiverzita obray Bro, 005