Application of computer image analysis for fast determination of size and shape of potato tubers

Pełen tekst

(1)Acta Agrophysica, 2007, 10(1), 47-57. ZASTOSOWANIE KOMPUTEROWEJ ANALIZY OBRAZU /(1,$ &,,.6=7$à78%8/:=,(01,$.$. '26=<%.,(*22.5( :,(/.2. Marek Gancarz, Krystyna Konstankiewicz ,QVW\WXW$JURIL]\NLLP%RKGDQD'REU]D. VNLHJR3$1XO'RZLDGF]DOQD 4, 20-290 Lublin. e-mail: marko@demeter.ipan.lublin.pl. S t r e s z c z e n i e . Praca przedstawia zastosowanie komputerowej analizy obrazu do szybkiego e. . X*\WR *FND*G]QLFK]GRNáDGQRFLJLPLHU]F ich UHdQLFH ] GRNáDGQRFL PP'RDNZL]\FMLREUD]yZZ\NRU]\VWDQRXU]G]HQLHRQLHSU]H(URF]\VW\FK FLDQDFK Z\SRVD*RQHZNDPHU &&'LXNáDG *DUyZHNRZLHWODMF\FKXPLHV]F]RQ\ZHZQWU]EDGaQ\ RELHNW 8U]G]HQLH XPR*OLZLD Z\NRQ\ZDQLH ]GM ü FDá\FK EXOZ Z VWDá\FK L NRQWURORZDQ\FK ZDUXQNDFK RSW\F]Q\FK $QDOL] RWU]\PDQ\FK REUD]yZEXOZ]LHPQLDNDGRNRQDQRSU]\Z\NRU]\VWaQLX VHNZHQFML SU]HNV]WDáFH RSDUW\FK QD ILOWURZDQLX L ELQDU\]DFML ] RIHUW\ SURJUDPX $SKHOLRQ Zastosowanie opracowanej SURFHGXU\ SR]ZDOD QD V]\ENLH RNUHOHQLH ZLHONRFL L NV]WDáWX EXOZ\ SRGF]DV MHGQHJR SRPLDUX GOD ZLHOX EXOZ MHGQRF]HQLH :LHONRü EXOZ\ RNUHODáD SRZLHU]FKQLD przypadajFDQDSáDVNLREUD]EXOZ\DMHMNV]WDáW–Z\GáX*HQLHRWU]\PDQHJRREUD]XEXOZ\0R*OLZH jHVW UyZQLH* X*\FLH SRZ\*V]HM PHWRG\ GR V]\ENLHM RFHQ\ EXOZ Z SUDNW\FH ODERUDWRU\MQHM MDN UywQLH*NRnWUROLMDNRFLLMHGQRURGQRFLVXURZFD. RNUH O QLD ZLHONR FL L NV]WDáWX NLONXG]LHVL FLX EXOZ ]LHPQLDND MHGQRF]H QLH 'R EDGD. *. EXOZRUy Q\FKUR]PLDUDFKLNV]WDáFLHZD. 6áRZDNOXF]RZH. : analiza obrazu, morfologia bulwy ziemniaka. 3. :67. *. . . do lepszego wykorzy u-. 3URGXFHQFL \ZQR FLFRUD]ZL NV] XZDJ SU]\NáDGDM . . VWDQLD UROQLF]\FK VXURZFyZ UR OLQQ\FK RUD] SRSUDZ\ LFK MDNR FL 2 MDNR FL V. . . URZFD L SU]\GDWQR FL SU]HWZyUF]HM GHF\GXM PLQ FHFK\ PRUIRORJLF]QH .RORU. . . aczenie przy wyborze przez konsumentów danego produktu, oraz jego zastosowaQLX Z SU]HP\OH Achenbach 2001, Abbott 1999, 'REU]DVNL L 5\EF]\VNL 2000, ZLHWOLNRZVND 6WXGPDQ

(2) 'ODWHJR ZD*QD MHVW XSUDZD. M GUQR üNV]WDáWZLHONR üVPDNWRFHFK\PRUIRORJLF]QHPDM FHGHF\GXM FH]Q.

(3) M. GANCARZ, K. KONSTANKIEWICZ. 48. . *. k-. WDNLFKRGPLDQNWyUH]DFKRZDM SRZ\ V]HFHFK\SRGF]DVFDáHJRSURFHVXSURGX. *. FMLMDNUyZQLH SRGF]DVSU]HWZDU]DQLD. . . 3UDFH EDGDZF]H SRWZLHUG]DM ZSá\Z ZLHONR FL NV]WDáWX RZRFyZ L ZDU]\Z QD. . SRZVWDZDQLHXV]NRG]H ]DUyZQR]HZQ WU]Q\FKMDNLZHZQ WU]Q\FKNWyU\FKGHWHNFMD. *. *. zrokowej do termograficznej (Baritelle i Hyde 1999, Baranowski i in. 2005, Mohsenin 1986, Diener i in. 1979, Studman 2001). W GRW\FKF]DVRZ\FKEDGDQLDFKZ\ND]DQR *HPLNURVWUXNWXUDWNDQNLPL NLV]RZHMMHVW FHFK FKDUDNWHU\VW\F]Q RELHNWX ]DOH*\ RG PLHMVFD Z obiekcie (Gancarz i in., 2005, Konstankiewicz i in. 2002, Zdunek i Konstankiewicz 2004). MHVW PR OLZD Uy Q\PL PHWRGDPL RG Z. . . -HGQ ]SRZV]HFKQLHXSUDZLDQ\FKUR OLQZ3ROVFHMHVW]LHPQLDND3ROVNDMHVW. mniaka do przetwyUVWZD GHF\GXMH ZLHOH F]\QQLNyZ PL G]\ LQQ\PL QLHNWyUH FHFK\ PRUIRORJLF]QH WDNLH MDN ZLHONRü L NV]WDáW F]\ NRORU PL*V]X Baritelle i Hyde, 1999, Noordam i in. 2000)3URGXFHQFLIU\WHNZ\ELHUDMRGPLDQ\NWyU\FKEXOZ\. MHGQ\P ] MHM QDMZL NV]\FK SURGXFHQWyZ QD ZLHFLH 2 SU]\GDWQR FL EXOZ ]LH. . * ZLHWOLNRZVND

(4) . 0DVD EXOZ\ PR*H RNUHODü MHM ZLHONRü D UHGQLFH EXOZ\ UHGQLFD PDNVy-. PDM NV]WDáW Z\GáX RQ\ QDWRPLDVW SURGXFHQFL FKLSVyZ EXOZ\ RNU JáH /LVL VND . Baritelle i Hyde 1999, Tabatabaeefar 2002, ZLHWOLNRZVND

(5) . Pomiary mDV\MDNWH* UHGQLFGRNRQ\ZDQHSU]\X*y-. PDOQD L SURVWRSDGáD GR QLHM

(6) MHM NV]WDáW . . *. . FLX WUDG\F\MQ\FK PHWRG L XU] G]H V MHGQDN F]\QQR FLDPL SUDFRFKáRQQ\PL L. – odmian, bulw – SU]\GDWQH VWDM VL PHWRG\ V]\ENLH L QLHZ\PDJDMFH GX*\FK QDNáadów pracy (Hartmut 1997, Eguchi i in. 1997, Studman 2001

(7) &RUD] F] FLHM GR V]\ENLHMLRELHNW\ZQHMRFHQ\LORFLRZHMLMDNRFLRZHMPDWHULDáyZVWRVXMHVL DQaOL] REUD]X NWyUD QLH Z\PDJD GX*\FK QDNáDGyZ F]DVX DQL SUDF\ D SR]ZDODX]yVNDü GRNáDGQH Z\QLNL MHGQDN PHWRG\ WH SR]ZDODM QD Z\NRQDQLH SRPLDUyZ QDjF]DVRFKáRQQ\PL 6]F]HJyOQLH SU]\ GX \FK SDUWLDFK PDWHULDáX EDGDZF]HJR. FLHM GOD SRMHG\QF]HJR RELHNWX +DUWPXW (JXFKLLLQ+HLQHPDQQ. F]. i LQ

(8) .RQLHF]QH MHVW ZL F RSUDFRZDQLH V]\ENLHM L RELHNW\ZQHM PHWRG\ Zy]QDF]HQLDZLHONRFLLNV]WDáWXEXOZNWyUDPR*HPLHüSUDNW\F]QH]DVWRVRZanie. &HOHP SUDF\ E\áR ]DVWRVRZDQLH NRPSXWHURZHM DQDOL]\ REUD]X GR V]\ENLHJR. i MHGQRF]HVQHJRZ\]QDF]HQLDZLHONRFLLNV]WDáWXNLONXG]LHVL FLXEXOZ]LHPQLDND 0$7(5,$à,0(72'<. *. . 'R EDGD X \WR EXOZ\ ]LHPQLDND RGPLDQ\ 3DVDW SRFKRG] FH ]H ]ELRUX SU]HSURZDG]RQHJR Z =DNáDG]LH 3U]HWZyUVWZD L 3U]HFKRZDOQLFWZD =LHPQLDND. Instytutu Hodowli i. . . a-. $NOLPDW\]DFML 5R OLQ Z -DGZLVLQLH 'R EDGD QDG ]DVWRVRZ. . . *. QLHP NRPSXWHURZHM DQDOL]\ REUD]X GR RNUH OHQLD ZLHONR FL L NV]WDáWX X \WR EXOZ. bez wiGRF]Q\FKXV]NRG]H]HZQ WU]Q\FKZD*FND*G]QLFK]GRNáDGQRFLJ RUD]PLHU]FUHGQLFHEXOZ\VXZPLDUN]GRNáDGQRFLPPZ celu wyznaczenia.

(9) ZASTOSOWANIE KOMPUTEROWEJ ANALIZY OBRAZU. . . . (. 49. i-. ZLHONR FL L NV]WDáWX EXOZ\ PHWRG WUDG\F\MQ L Sy QLHMV]HJR SRUyZQDQLD Z\Q. . y. *. NyZ .V]WDáW EXOZ\ RNUH ODQ\ MDNR ZVSyáF]\QQLN Z GáX HQLD Z\OLF]RQR ZHGáXJ Z]RUX. ZLHWOLNRZVND

(10) . TS = a/b , gdzie: TS – ZVSyáF]\QQLN Z\GáX*HQLD E – UHGQLFD EXOZ\ PLHU]RQD Z]GáX* RVL VWolon-ZLHU]FKRáHND–QDMZL NV]DUHGQLFDEXOZ\PLHU]RQDSURVWRSDGOHGRWHMRVL Do akwizycML REUD]yZ VNRQVWUXRZDQR V]HFLHQQ NRPRU R QLHSU]H(URF]\VW\FK FLDQDFKRGáXJRFLFP:FHQWUXPJyUQHMSRGVWDZ\QDZ\VRNRFLFPXPLHVzF]RQRNDPHU &&'..Z\SRVD*RQZRELHNW\Z7UHYLGRQ1DSU]eNWQ\FK JyUQHM SRGVWDZ\ZRGOHJáRFLFPRGND*GHJRZLHUzFKRáND]DPRQWRZDQR *DUyZNL W\SX 9 : *8 ; ]DVLODQH QaSL FLHP 9 ]DVLODF]HP =0 (8PR*OLZLDRQSá\QQUHJXODFM QDSL FLDZ]DNUHVLH-9XPR*OLZLDMF GREyU RSW\PDOQHJR RZLHWOHQLD Z FHOX X]\VNDQLD REUD]X EDGDQHJR obiektu bez wiGRF]QHJR FLHQLD U\V

(11) :V]\VWNLH FLDQ\ ZHZQWU] NRPRU\ SRNU\WR ELDá\P SDSLeUHPUR]SUDV]DMF\PZLDWáR ELEXáDMDNoFLRZDUHGQLDILUP\67$1'$5'

(12) *. 7DN ]EXGRZDQD NRPRUD XPR OLZLD X]\VNDQLH REUD]X EDGDQHJR RELHNWX EH]. widocznego jego cieniaDE\ZDQDOL]LHNRPSXWHURZHMREUD]XRWU]\PDüRELHNW\R jednoznacznych granicach. =DVWRVRZDQLH ZLHORSXQNWRZHJR JyUQHJR RZLHWOHQLD GDMH PR*OLZRü ]DVWo*. . . VRZDQLD NRPRU\ GR Uy Q\FK FHOyZ QS RNUH OHQLD XV]NRG]H QD SRZLHU]FKQL. obiektu (nie tylko ziemniaka).. /. /. &. & 2 a). b). Rys. 1. Schemat budowy komory do akwizycji obrazów: a) widok z góry, b) widok z boku: L – RZLHWOHQLH2– badany obiekt, C – kamera lub aparat fotograficzny Fig. 1. Schematic of chamber for image acquisition, a) view from top, b) view from side, L – lights, O – sample object, C – CCD camera or photo camera.

(13) 50. M. GANCARZ, K. KONSTANKIEWICZ. . oa-. %DGDQ\ RELHNW XPLHV]F]DQRQDGROQHMSRGVWDZLHV]H FLHQQHMNRPRU\SURVW. . *. SDGáHMGRRVLNDPHU\3U]\SRPRF\WHJRXU] G]HQLDPR OLZDMHVWDNZL]\FMDREU. ]yZ Z VWDá\FK L NRQWURORZDQ\FK ZDUXQNDFK RSW\F]Q\FK FR MHVW QLH]E GQH Z. . . NRPSXWHURZHMLOR FLRZHMLMDNR FLRZHMDQDOL]LHRWU]\PDQ\FK]GM ü. Do analizy otrzymanych obrazów wykorzystano program Aphelion (Gancarz i in. 2007, Wojnar i in. 2002) XPR*OLZLDMF\ SU]\ ]DVWRVRZDQLX RGSRZLHGQLFK SU]eNV]WDáFH Z\]QDF]HQLH SDUDPHWUyZ RSLVXMF\FK ZLHONRü L NV]WDáW SáDVNLHJR Rb . UD]X EXOZ\ 2WU]\PDQR GZD SDUDPHWU\ OLF]ED SLNVHOL SU]\SDGDM F QD SáDVNL. obraz bulwy – ARUD]Z\GáX*enie – E (rys. 2).. D. E. (. I. DE DE. II. III. erem 10, I – – HIHNW DQDOL]\ SROH RJUDQLF]RQH NU]\Z RNUHOD OLF]ED SLNVHOL A ZLHONRü bulwy), III –Z\GáX*HQLDEXOZ\ E NV]WDáW

(14) DLE–UHGQLFHHOLSV\ZSLVDQHMZREUD]EXlwy Fig. 1. Example of determination of size and shape of the potato tuber tissue number 10, I – flat tuber image, II – result of analysis: the area limited by curvilinear boundary defines the number of pixels A (tuber size), III – tuber elongation (tuber shape), a and b – ellipse diameters inscribed in tuber image. Rys. 2.. 3U]\NáDGRZH RNUH OHQLH ZLHONR FL L NV]WDáWX EXOZ\ ]LHPQLDND R]QDF]RQHM QXP. SáDVNL REUD] EXOZ\ ,,. o-. $E\ RWU]\PDü SRZLHU]FKQL SáDVNLFK REUD]yZ EXOZ Z MHGQRVWNDFK 6, Z\N. QDQR ZNRPRU]H]GM FLHLGHQW\F]Q\FKSáDVNLFKRELHNWyZR]QDQHMSRZLHU]FKQL. przy takich samych ustawieniach jak w przypadku bulw ziemniaka. Umieszczono je na dolnej podstawie komory: jeden w osi kamery, drugi jak najdalej od osi . NDPHU\ DOH Z SROX MHM ZLG]HQLD D WU]HFL Z SRáRZLH RGOHJáR FL SRPL G]\ QLPL. . *. :\NRQDQR WR Z FHOX RNUH OHQLD Eá GX MDNL PR HP\ SRSHáQLü SRGF]DV SRPLDUyZ. *. . . Skalowanie. RELHNWyZ XPLHV]F]RQ\FK Z Uy Q\FK RGOHJáR FLDFK RG RVL NDPHU\. . . GRNRQDQHQDRELHNFLHXPLHV]F]RQ\PZSRáRZLHRGOHJáR FLPL G]\RVL NDPHU\D QDMGDOV]\P SXQNWHP RGGDORQ\P RG WHM RVL SR]ZROLáR Z\]QDF]\ü PDNV\PDOQ\. . *. . Eá G MDNL PR HP\ SRSHáQLü SRGF]DV RNUH ODQLD SRZLHU]FKQL SáDVNLHJR REUD]X. bulwy, który nie przekracza 1,5%..

(15) ZASTOSOWANIE KOMPUTEROWEJ ANALIZY OBRAZU. 51. WYNIKI I DYSKUSJA . *. . *. :SLHUZV]HMNROHMQR FL]ZD RQRL]PLHU]RQR UHGQLFHND GHM]EXOZDE\. . . RNUH OLüMHMZLHONR üLNV]WDáW WDE

(16) . *. . 1DVW SQLH SU]\ X \FLX VNRQVWUXRZDQHJR XU] G]HQLD U\V

(17) Z\NRQDQR ]GM . . FLHQLHVW\NDM F\FKVL EXOZNWyUHE\áRSRGGDQHDQDOL]LHNRPSXWHURZHMZFHOX. . *. a-. MHGQRF]HVQHJRZ\]QDF]HQLDZLHONR FLLNV]WDáWXND GHMEXOZ\ZLGRF]QHMQDREU. zie (rys. 3).. *. /LF]E EXOZ GHWHUPLQRZDá\ Z\PLDU\ NRPRU\ SDUDPHWU\ WHFKQLF]QH X \WHM. . NDPHU\RUD]ZLHONR üEXOZ. dokonanie analizy uzyskanego obrazu w celu wyzna eURZHM DQDOL]\ GRNRQ\ZDQR Z\NRU]\VWXMF VHNZHQFM SU]HNV]WDáFH RSLHUDMcych VL QDILOWURZDQLXLELQDU\]DFMLZ\NRU]\VWXMFSURJUDP$phelion. .ROHMQ\PNURNLHPE\áR. F]HQLD SDUDPHWUyZ REUD]X EXOZ\ RNUH ODM F\FK MHM ZLHONR ü L NV]WDáW .RPSXW. Rys. 3. =GM FLHEXOZZ\NRQDQHZFHOXRNUHOHQLDLFKZLHONRFLRUD]NV]WDáWX Fig. 3. Photo of 30 potato tubers for determination of their size and shape. iono obrazy w odFLHQLDFKV]DURFLGODWU]HFKVNáDGRZ\FK5 – czerwony, G – zielony i B – niebieski. Wybrano z nich najbardziej kontrastowy B – niebieski w celu dalszej analizy : SLHUZV]HM ID]LH DQDOL]\ ] REUD]X NRORURZHJR Z\RGU EQ. ]DVWRVRZDQRILOWUPHGLDQRZ\[SLNVHODDE\Z\HOLPLQRZDüHZHQWXDOQHV]XP\.

(18) 52. M. GANCARZ, K. KONSTANKIEWICZ. Nast SQLH SU]\ SRPRF\ ELQDU\]DFML ] JyUQ\P SURJLHP Z\]QDF]RQR JUDQLFH. u-. SáDVNLFK REUD]yZ EXOZ QD ]GM FLX .ROHMQ\P NURNLHP E\áR Z\SHáQLHQLH HZHQW. . DOQ\FK RWZRUyZ ZHZQ WU] REUD]X EXOZ\ L HW\NLHWRZDQLH REUD]X SU]\SLVDQLH. kolejnym znalezionym obszarom unikalnych identyfikatorów - kolor). W dalszej fazie analizy dokonano konwersji obrazu zetykietowanego na zestaw obszarów w FHOXZ\NRQDQLDSRPLDUyZ U\V

(19) 2VWDWQLPNURNLHPDQDOL]\E\áSRPLDUSáDVNLe*. JRREUD]XND GHMEXOZ\. aa-. :Z\QLNXSU]HSURZDG]RQHMDQDOL]\NRPSXWHURZHMRWU]\PDQRSDUDPHWU\Sá. *. VNLHJRREUD]XND GHMEXOZ\ U\V

(20) -HGHQ]SDUDPHWUyZWROLF]EDSLNVHOLSU]\S. – ADGUXJLWRZ\GáX*HQLHSáDVNLHJRREUD]XEXOZ\ – E. Po przeprowadzeniu skalowania parametru A RWU]\PDQR SRZLHU]FKQL Sáaskiego obrazu bulwy S w cm2. . GDM F QDSáDVNLREUD]EXOZ\. Rys. 4. (IHNW DQDOL]\ NRPSXWHURZHM XPR*OLZLDMF\ X]\VNDQLH OLF]E\ SLNVHOL SU]\SDGDMcych na SáDVNLREUD]ND*GHMEXOZ\RUD]Z\]QDF]HQLHNV]WDáWXEXOwy Fig. 4. Result of computer analysis allowing the determination of the number of pixels for. the flat image of each tuber and definition of tuber shape . . . : WDEHOL ]HVWDZLRQR X]\VNDQH ZLHONR FL RNUH ODM FH EXOZ SRZLHU]FKQL . askiego obrazu bulwy S, maV W Z\GáX*HQLH E SU]\ X*\FLX DQDOL]\ REUD]X L ZVSyáF]\QQLNZ\GáX*HQLDZ\]QDF]RQ\PHWRGWUDG\F\MQTS.. Sá.

(21) ZASTOSOWANIE KOMPUTEROWEJ ANALIZY OBRAZU. 53. Tabela 1.=HVWDZLHQLHZ\QLNyZRNUHOHQLDZLHONRFLLNV]WDáWXEXOZRGPLDQ\3DVDWX]\VNDQ\FK przy pomocy komputerowej analizy obrazu oraz metodami poZV]HFKQLHX*\ZDQ\PL Table 1. Comparison of size and shape results of the 30 tubers of variety Pasat obtained from computer image analysis and with traditional methods TN 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30. S (cm2) 76,97 62,29 49,11 37,45 35,95 37,09 45,59 35,69 34,24 35,28 42,26 38,31 38,84 45,92 36,88 50,74 38,37 44,12 41,55 41,41 37,61 43,47 34,94 47,28 31,41 51,83 48,90 41,29 64,67 48,93. W (g) 198,47 164,80 119,30 87,18 78,44 75,13 102,65 86,58 87,79 84,22 109,79 82,52 70,65 92,70 95,24 141,44 96,09 103,93 95,27 91,45 93,20 113,75 94,00 135,27 71,73 143,86 120,66 96,91 158,57 120,46. E 0,283 0,025 0,044 0,053 0,163 0,047 0,094 0,160 0,200 0,072 0,131 0,137 0,050 0,050 0,085 0,018 0,090 0,087 0,173 0,091 0,288 0,096 0,109 0,021 0,093 0,050 0,202 0,043 0,256 0,140. TS 0,752 0,997 0,967 0,969 0,905 0,979 0,943 0,845 0,811 0,962 0,913 0,930 0,967 1,000 0,952 0,987 0,927 0,933 0,842 0,979 0,771 0,925 0,940 0,984 0,971 0,986 0,850 0,960 0,790 0,893. TN – numer bulwy, S -SRZLHU]FKQLDSáDVNLHJRREUD]XEXOZ\ ZLHONRüEXOZ\

(22) :– masa bulwy, E –Z\GáX*HQLH NV]WDáWEXOZ\

(23) 76–ZVSyáF]\QQLNZ\GáX*HQLDREOLF]RQ\WUDG\F\MQLH NV]WDáWEXOZ\

(24) TN – tuber number, S - area of flat tuber image (tuber size), W – mass of tuber, E – tuber elongation (tuber shape), TS – elongation ratio of tuber obtained with traditional method (tuber shape)..

(25) 54. M. GANCARZ, K. KONSTANKIEWICZ. W wyniku przeprowadzonego eksperymentu dla bulw odmiany Pasat otrzy*. . . . PDQR ]DOH QR ü SRPL G]\ SRZLHU]FKQL SU]\SDGDM F QD SáDVNL REUD] EXOZ\ L. . *. d a-. PDV WHMEXOZ\ U\V

(26) 3RUyZQDQRZVSyáF]\QQLNZ\GáX HQLDZ\]QDF]RQ\MH Q . *. *. ] SRZV]HFKQLH X \ZDQ\FK PHWRG L Z\GáX HQLH SáDVNLHJR REUD]X EXOZ\ Z\]Q. czonej przy pomocy komputerowej analizy obrazu (rys. 6). . . 5 . . . . 6>PP @. . . . . :>J@. Rys. 5.:\NUHV]DOH*QRFLSRZLHU]FKQLSáDVNLHJRREUD]XEXOZ\– S od masy bulwy – W Fig. 5. Diagram of dependence between area of flat tuber image – S and tuber mass – W . . 5 . . . 76. . . . . . . . . . . . . . (. Rys. 6. 3RUyZQDQLH ZVSyáF]\QQLND Z\GáX*HQLD Z\]QDF]RQHJR WUDG\F\MQLH – 76 L Z\GáX*HQLD Sáaskiego obrazu bulwy wyznaczone przy pomocy komputerowej analizy obrazu – E Fig. 6. Comparison of elongation ratio obtained with traditional method – TS and elongation of flat tuber image obtained from computer image analysis – E.

(27) ZASTOSOWANIE KOMPUTEROWEJ ANALIZY OBRAZU. 55. 2. = 0, SRPL G]\ SRZLHU]FKQL Sáa * aF]HQLDZLHONRFLEXOZ\0R*OLZHMHVWZL FEDUG]RV]\ENLHLGRNáDGQHZ\]QDF]eQLH ZLHONRFL NLONXG]LHVL FLX EXOZ MHGQRF]HQLH FR ]QDF]QLH RJUDQLF]D F]DVoFKáRQQRüMDNUyZQLH*QDNáDG\SUDF\ 1DU\VXQNXSU]HGVWDZLRQR]DOH*QRüSRPL G]\ZVSyáF]\QQLNLHPZ\GáX*eQLD Z\]QDF]RQHJR WUDG\F\MQLH L Z\GáX*HQLD SáDVNLHJR REUD]X EXOZ\ Z\]QDF]o:\VRNL ZVSyáF]\QQLN GHWHUPLQDFML 5. VNLHJR REUD]X EXOZ\ D PDV EXOZ\ SR]ZDOD QD X \FLH WHMPHWRG\ZFHOXZ\]Q. QHJRSU]\SRPRF\NRPSXWHURZHMDQDOL]\REUD]X:VSyáF]\QQLNGHWHUPLQDFMLMHVW. bardzo wysoki R2 = Z\NUHVX Z\QLND *H PR*OLZH MHVW ]DVWRVRZDQLH WHM meWRG\ WDN*H GR RNUHODQLD NV]WDáWX EXOZ\ 7DN MDN Z SU]\SDGNX Z\]QDF]HQLD . *. *. ZLHONR FL EXOZ\ WX UyZQLH PR HP\ Z\]QDF]\ü NV]WDáW NLONXG]LHVL FLX EXOZ. jednoF]HQLH]QDF]QLHVNUDFDMF czas pracy. . 2WU]\PDQH Z\QLNL SR]ZDODM VWZLHUG]Lü. . . *H ZLHONRü EXOZ\ EDGDQHM RGPLa-. *. oZLHU]FKQL SáDVNLHJR REUD]X EXOZ\ SU]\ SRPRF\ NRPSXWHURZHM DQDOL]\ REUazu.. Q\ NWyU RNUH OD MHM PDVD PR H E\ü Z SURVW\ L V]\ENL VSRVyE RSLVDQD SU]H] S. . *. . *. -HGQRF]H QLH PR QD RNUH OLü MHM NV]WDáW R]QDF]RQ\ SU]H] Z\GáX HQLH 0HWRGD WD. *. PR H E\ü ]DVWRVRZDQD GR ZL NV]HM OLF]E\ EXOZ FR ]QDF]QLH VNUDFD F]DV L QLH. *. *. . Z\PDJDGX \FKQDNáDGyZSUDF\3R]ZDODWH QDV]\ENLHLRELHNW\ZQHRNUH OHQLH. . *. *. ZLHONR FL L NV]WDáWX EXOZ\ 0R H E\ü RQD UyZQLH SRPRFQD Z GRERU]H EXOZ WHM. . . VDPHMZLHONR FLLNV]WDáFLHGREDGD . WNIOSEK. . oeprowadzone pomiary dla odmiany PaVDW Z\ND]Dá\ GX* ]JRGQRü SRPL G]\ SarameWUDPL SRZLHU]FKQL L Z\GáX*HQLD SáDVNLHJR REUD]X EXOZ\ RNUHODMF\PL MHM wielNRüRUD]NV]WDáW]Z\QLNDPLX]\VNDQ\PLPHWRGDPLWUDG\F\MQ\PL0HWRGDPR*H E\ü ]DVWRVRZDQD GR V]\ENLHM RFHQ\ ZLHONRFL L NV]WDáWX EXOZ EDGanej odmiany w =DSURSRQRZDQR PHWRG NRPSXWHURZHM DQDOL]\ REUD]X GR V]\ENLHJR MHGQ. . F]HVQHJR RNUH OHQLD ZLHONR FL L NV]WDáWX NLONXG]LHVL FLX EXOZ ]LHPQLDND 3U]. *. . . SUDNW\FHODERUDWRU\MQHMMDNUyZQLH NRQWUROLMDNR FLLMHGQRURGQR FLVXURZFD. 0,(11,&7:2. 3,. Abbott J.A., 1999. Quality measurement of fruits and vegetables. Postharvest Biology and Technology 15, 207-225. Achenbach J., 2001. The power of light. National Geographic, 10, 2 -29. %DUDQRZVNL3/LSHFNL-0D]XUHN::DOF]DN57'HWHNFMDXV]NRG]HPHFKDQLFznych MDEáHN]Z\NRU]\VWDQLHPWHUPRJUDILLActa Agrophysica, 125, 6(1), 19-29. Baritelle, A.L., Hyde G.M., 1999. The effect of tuber size on the impact sensitivity of potato tubers. Transaction of the ASAE, 42(1), 159-162..

(28) 56. M. GANCARZ, K. KONSTANKIEWICZ. Diener R.G., Elliott K.C., Nesselroad P.E., Ingle M., Adams R.E., Blizzard S.H., 1979. Bruise Energy of Peaches and Apples. Transaction of the ASAE, 22(2), 287-290. DoEU]DVNL MU % 5\EF]\VNL 5 ,QWHUSUHWDFMD IL]\F]QD RFHQ\ EDUZ\ Z ]DVWRVRZDQLX GR klasyfiNDFMLMDNRFLRZHMMDEáHN$Fta Agrophysica, 37, 17-27. 'REU]D. VNL MU % 5\EF]\VNL 5 &XVWRPHU SHUFHSWLRQ RI FRORU RQ 3ROLVK PDUNHW RI IUXLW. and vegetable. Information & technology for sustainable & vegetable production. Post Harvest Technology, 25, 1-8. Eguchi T., Kitano M., Eguchi H., 1997. On-line system for volume measurement in sweet potato tuber. Biotronics, 26, 103-106. Gancarz M., KonstankiewiF].3DZODN.=PLHQQRüSDUDPHWUyZVWUXNWXU\NRPyUNRZHM w bulwie ziemniaka. (Variability of parameters of cellular structure in potato tubers), Acta Agrophysica, 127, 6(3), 625-638. Gancarz M., Konstankiewicz K., Pawlak K., Zdunek A., 2007. Analysis of plant tissue images obtained by confocal tandem scanning reflected light microscope. Int. Agrophysics, 21, 1, 49-54. Hartmut Gall., 1997. A ring sensor system using a modified polar coordinate system to describe the shape of irregular objects. Meas. Sci. Technol., 8, 1228-1235. Heinemann P.H., Pethare N.P., Morrow C.T., 1996. An automated inspection station for machinevision grading of potatoes. Machine Vision and Applications, 9 (1), 14-19. Konstankiewicz K., Czachor H., Gancarz M., Król A., Pawlak K., Zdunek A., 2002. Cell structural parameters of potato tuber tissue. Int. Agrophysics, 16, 2. /LVLVND*=LHPQLDNMDNRVXURZLHFGODSU]HP\VáX3RVW1DXN5ROQ-39. Mohsenin N.N., 1986. Physical properties of plant and animal materials. Gordon and Breach Science publishers, New York. Noordam J.C., Otten G.W., Timmermans A.J.M., van Zwol B.H., 2000. High-speed potato grading and quality inspection based on a color vision system. Proc. SPIE Vol. 3966, p. 206-217, Machine Vision Applications in Industrial Inspection VIII, Kenneth W. Tobin; Ed. Studman C. J., 2001. Computers and electronics in postharvest technology – a review. Computers and electronics in Agriculture, 30, 109-124. ZLHWOLNRZVND. UHG

(29) 6XURZFHVSR*\ZF]HSRFKRG]HQLDUROLQQHJR6**:. Tabatabaeefar A., 2002. Size and shape of potato tubers. Int. Agrophysics, 16, 4, 301-305. :RMQDU / .XU]\GáRZVNL . - 6]DOD - 3UDNW\ND DQDOL]\ REUD]X 3ROVNLH 7RZDU]\VWZR. Stereologiczne, Kraków. Zdunek A., Konstankiewicz K., 2004. Acoustic emission in investigation of plant tissue microcracking. Transactions of ASAE, 47 (4), 1171-1177..

(30) ZASTOSOWANIE KOMPUTEROWEJ ANALIZY OBRAZU. 57. APPLICATION OF COMPUTER IMAGE ANALYSIS FOR FAST DETERMINATION OF SIZE AND SHAPE OF POTATO TUBERS Marek Gancarz, Krystyna Konstankiewicz Institute of Agrophysics, Polish Academy Sciences, uO'RZLDGF]DOQD-290 Lublin e-mail: marko@demeter.ipan.lublin.pl A b s t r a c t . The paper presents the application of computer image analysis for fast determination of size and shape of potato tubers. In the experiment 30 tubers with different sizes and shapes were tested. Each tuber was weighed with 0.01g accuracy and its diameter was measured with 0.1 mm accuracy. Images of potato tubers were acquired by means of a device consisting of a box chamber with opaque walls, a CCD camera, and a set of bulbs illuminating objects placed inside the chamber for examination. This device permitted photos of the tubers to be taken under conditions of controlled optical parameters. Analysis of obtained images was performed using a set of options of the Aphelion software package. The developed procedure allows fast determination of tuber size and shape during one measurement for many tubers simultaneously. The size of a tuber is described by the area of flat tuber image, while its shape is defined by the elongation of tuber image. The method can be applied for fast estimation of potato tubers in laboratory practice and also for the control of quality and homogeneity of material. K e y w o r d s: image analysis, potato tuber morphology.

(31)