Seed Size and Shape Analysis of Registered Common Bean ( Phaseolus vulgaris L . ) Cultivars in Turkey Using Digital Photography

In this study, the size and shape properties of twelve different bean cultivars (eleven varieties and a genotype) were determined using image processing method on digital photography and the cultivars were grouped depending on their properties. The highest mean values of geometric diameter, projected area, length, width and shape factor in the study were determined in the ‘Çayırlı’ genotype. ‘Terzibaba’ cultivar, more spherical shape at width orientation compared to other cultivars, had the lowest value of thickness (4.6 mm) and more circular at the width orientation among the cultivars. The highest mean of thickness (6.0 mm) were measured in ‘Elkoca-05’ and ‘Noyanbey-98’ cultivars. The sphericity mean of ‘Karacaşehir-90’ (70%) was found higher than those of the other cultivars, while the geometric mean diameter, projected area, length and width values had the lowest mean. ’Yunus-90’ and ‘Göynük-98’ cultivars had the lowest sphericity mean values of 58.9% and 58.7%, respectively. According to the results of Principal Component Analysis (PCA), the most important variables of the first component were mass, geometric mean diameter, surface area, volume, projected area, equivalent diameter, perimeter and length. According to the physical properties of bean cultivars ‘Aras-98’, ‘Elkoca-05’, ‘Göynük-98’, ‘Noyanbey-98’, ‘Yakutiye-98’ and ‘Yunus-90’ were oblongus; ‘Karacaşehir-90’ and ‘Önceler-98’ were ellipticus; ‘Akman-98’, ‘Kantar-05’ and ‘Terzibaba’ were subcompressus and ‘Çayırlı’ genotype was compressus. The size and shape features of the bean cultivars are considerably crucial parameters in terms of designing the seed metering mechanism of seed drills, and transportation, separating and sizing systems. So, the size and shape data of bean cultivars are of mostly importance to engineers, machine manufacturers and machine designers.


221
and shape index.Sayinci et al (2012) determined the size and shape of some orange cultivars.Boydas et al (2012) measured the physical properties of loquat (Eriobotrya japonica [Thunb.Lindl.])cultivars using image process and classical methods.Yurtlu & Yeşiloğlu (2011) determined the physical and mechanical properties of chestnuts.
Information on the size and shape features of bean cultivars such as their dimensions (length, width and thickness), projected area, mass, volume, elongation at different orientations and sphericity is considerably important in the design of seed metering mechanisms of seed drills, separating, sizing and handling systems.Therefore, the objective of the study was to determine the size and shape of 11 registered common bean cultivars and a genotype grown in Turkey.An image processing system, which is an effective method for automated assessments, was used in order to determine the physical properties of bean.

Material and Methods
In this study, 11 registered common bean cultivars 'Terzibaba', and a genotype ('Çayırlı') were studied (Figure 1).In order to determine the size and shape properties of common bean cultivars, image-processing method was used.The digital photographs of seeds were taken in a dark room by a Nikon D300 digital SLR camera.The camera was fixed on a frame in a position perpendicular to a fiberglass surface at fixed height of 35 cm.The seeds were placed on the transparent fiberglass plate in the matrix form of 2×5 at two orientations, namely width and thickness orientations (Figure 2).In order to position seeds on the fiberglass with dimensions of 10×7 cm in two orientations, putty was used.The width orientation represents the natural position of the seeds on horizontal surface.Therefore, the digital images taken in two orientations of the seeds were possible to determine the size and shape properties.The color of fiberglass was selected opposite to the color of seeds.A ruler with intervals of 1 mm was

223
placed beside the fiberglass plate to converse pixel to mm.The unit conversion ratio of the images was 46.2 pixels mm -1 .A cable release was used to prevent the vibration during the shooting.In order to capture an image without shadow on the surface, three paraflashes were positioned in three different directions (Figure 3).Two softboxes were mounted on two paraflashes and placed over the fiberglass plate with an angle of 45 o .The third paraflash was placed under the fiberglass plate with a distance of 20 cm.
The light intensity was adjusted by an exposure meter.The light intensity of the paraflash under the fiberglass plate was higher than those of the paraflashes above the fiberglass plate.SigmaScan ® Pro 5.0 software was used to process the images.The pixel value corresponding to the 100 mm distance was used for calibration in the photo images.The projected area (PA, mm 2 ), perimeter (P, mm), equivalent diameter, major and minor axis lengths and shape factor (SF) were determined for each seed at width and thickness orientation using SigmaScan ® Pro 5.0 software.The major length measured at the width orientation means the length (L, mm) of seed, the minor axis length measured at the width orientation means the width (W, mm) of seed, the minor axis length measured at the thickness orientation means thickness (T, mm) of seed.The major axis length measured at the width orientation was the same with one measured at the thickness orientation.The surface area (S, mm 2 ), elongation, geometric mean diameter (D g , mm) and sphericity (φ, %) were calculated for each seed using the following equations (L: length, mm; W: width, mm; T: thickness, mm) (Mohsenin 1986): thickness, mm) (Mohsenin 1986): (1) (2) Elongation, the ratio of major axis length to minor axis length (Fıratlıgil-Durmus et al 2010), was calculated for three orientations: Elongation at the width orientation (E w ); Elongation at the thickness orientation (E t ); Elongation at the vertical orientation (E v ); The cultivars were classified according to Şehrali (1988): ellipticus (ellipsoid and L/W = 1.51-1.71), (1) thickness, mm) (Mohsenin 1986): (1) (2) Elongation, the ratio of major axis length to minor axis length (Fıratlıgil-Durmus et al 2010), was calculated for three orientations: Elongation at the width orientation (E w ); Elongation at the thickness orientation (E t ); Elongation at the vertical orientation (E v ); The cultivars were classified according to Şehrali (1988): ellipticus (ellipsoid and L/W = 1.51-1.71),oblongus (long cylindroid, kidney shaped and L/W = 1.85-2.31),subcompressus (subcompressed, long Elongation, the ratio of major axis length to minor axis length (Fıratlıgil-Durmus et al 2010), was calculated for three orientations: Elongation at the width orientation (E w ); Elongation at the thickness orientation (E t ); Elongation at the vertical orientation (E v ); The cultivars were classified according to Şehrali (1988): ellipticus (ellipsoid and L/W = 1.51-1.71),oblongus (long cylindroid, kidney shaped and L/W = 1.85-2.31),subcompressus (subcompressed, long and (W/T)×L = 1.29-2.08),and compressus (more compressed, broad and (W/T)×L= 2.17-3.51).The shape factor (SF), determined by the software automatically, was calculated using the following equation (PA: projected area, mm 2 ; P: perimeter, mm) (SigmaScan ® Pro 2004): The surface area of the seeds (S, mm 2 ) was calculated from the below equation (McCabe et al 1986; Olajide &Ade-Omowaye 1999): Elongation, the ratio of major axis length to minor axis leng calculated for three orientations: Elongation at the width orientatio Elongation at the thickness orientation (E t ); The cultivars were classified according to Şehrali (1988): elli oblongus (long cylindroid, kidney shaped and L/W = 1.85-2.31)and (W/T)×L = 1.29-2.08),and compressus (more compressed, shape factor (SF), determined by the software automatically, was (PA: projected area, mm 2 ; P: perimeter, mm) (SigmaScan ® Pro 200 The surface area of the seeds (S, mm 2 ) was calculated from th Olajide &Ade-Omowaye 1999): The calculated volume values of the seeds (V, cm 3 ) was determine Seed mass values were determined using an electronic balance displacement method was used to determine experimental volu moisture content of the seeds was determined using an oven set a conducted at the Biological Material and Image Processing Labor shape properties of 12 common bean cultivars, 200 seeds wer totaling 2400 seeds.A completely randomized design and SPSS st of variance with a 95% confidence level (P = 0.05) and Duncan' significant differences (IBM SPSS ® Statistics 2010).The data w Analysis (PCA) in order to test the variables of 12 common bea between multiple variables were determined and the common covariance matrix obtained from the mean values of the size and output data sets including Eigen values and vectors and principal c

Results and Discussion
The mean values of moisture content, seed mass, thousand seed bean cultivars investigated in the study are presented in Table 1. the cultivars varied in the range of 3.39-6.40%(dry basis), the 0.656 g, and thousand seed mass in the range of 180-656 g. 'Ça seed mass (0.656 g), followed by 'Elkoca-05' (0.462 g), 'Göynükg).The differences between the mean values of seed mass for th The calculated volume values of the seeds (V, cm 3 ) was determined using the below equation.
Elongation, the ratio of major axis length to minor axis leng calculated for three orientations: Elongation at the width orientatio Elongation at the thickness orientation (E t ); The cultivars were classified according to Şehrali (1988): elli oblongus (long cylindroid, kidney shaped and L/W = 1.85-2.31)and (W/T)×L = 1.29-2.08),and compressus (more compressed, shape factor (SF), determined by the software automatically, was (PA: projected area, mm 2 ; P: perimeter, mm) (SigmaScan ® Pro 200 The surface area of the seeds (S, mm 2 ) was calculated from th Olajide &Ade-Omowaye 1999): The calculated volume values of the seeds (V, cm 3 ) was determine Seed mass values were determined using an electronic balance displacement method was used to determine experimental volu moisture content of the seeds was determined using an oven set a conducted at the Biological Material and Image Processing Labor shape properties of 12 common bean cultivars, 200 seeds wer totaling 2400 seeds.A completely randomized design and SPSS st of variance with a 95% confidence level (P = 0.05) and Duncan' significant differences (IBM SPSS ® Statistics 2010).The data w Analysis (PCA) in order to test the variables of 12 common bea between multiple variables were determined and the common covariance matrix obtained from the mean values of the size and output data sets including Eigen values and vectors and principal c

Results and Discussion
The mean values of moisture content, seed mass, thousand seed bean cultivars investigated in the study are presented in Table 1. the cultivars varied in the range of 3.39-6.40%(dry basis), the 0.656 g, and thousand seed mass in the range of 180-656 g. 'Ça seed mass (0.656 g), followed by 'Elkoca-05' (0.462 g), 'Göynükg).The differences between the mean values of seed mass for th (8) Seed mass values were determined using an electronic balance reading to 0.001 g.The toluene (C 7 H 8 ) displacement method was used to determine experimental volume and true density of the seeds.The moisture content of the seeds was determined using an oven set at 105 °C for 24 hours.All the tests were conducted at the Biological Material and Image Processing Laboratory.In order to determine the size and shape properties of 12 common bean cultivars, 200 seeds were selected randomly for each cultivar, totaling 2400 seeds.A completely randomized design and SPSS statistical software were used for analysis of variance with a 95% confidence level (P = 0.05) and Duncan's multiple comparison test to determine significant differences (IBM SPSS ® Statistics 2010).The data was subjected to Principal Component Analysis (PCA) in order to test the variables of 12 common bean cultivars in groups.The relationships between multiple variables were determined and the common bean cultivars were grouped using a covariance matrix

Results and Discussion
The mean values of moisture content, seed mass, thousand seed mass, and true density of the common bean cultivars investigated in the study are presented in Table 1.The mean values of moisture content of the cultivars varied in the range of 3.39-6.40%(dry basis), the mean seed mass in the range of 0.180-0.656g, and thousand seed mass in the range of 180-656 g. 'Çayırlı' genotype had the highest value of seed mass (0.656 g), followed by 'Elkoca-05' (0.462 g), 'Göynük-98' (0.450 g), and 'Yakutiye-98' (0.447 g).The differences between the mean values of seed mass for these three cultivars were not statistically significant.'Karacaşehir-90' had the lowest seed mass value (0.180 g).In a study by Güvenç & Güngör (1996), thousand seed mass values of seven common bean cultivars varied between 209.1 and 467.6 g.Firatlıgil-Durmuş et al ( 2010) also determined the thousand seed mass values of seven common bean cultivars as between 171.7 and 613.2 g.In a previous study, the highest value of thousand seed mass was obtained as approximately 751.8 g (at the moisture content of 18.33%) with 'Barbunia' (Cetin 2007).The highest mean values of true density of the cultivars was obtained with 'Göynük-98' (1.305 g cm -3 ) and 'Karacaşehir-90' (1.317 g cm -3 ).The lowest true density among the cultivars belonged to 'Önceler-98' with the value of 1.309 g cm -3 .This value is similar to that reported by Fıratlıgil-Durmuş et al ( 2010) approximately 1.220 g cm -3 for 'White speckled red bean' (from Czech Republic).Mean value of true density of 'Kantar-05' (1.279 g cm -3 ) is almost the same as that reported by Ozturk et al (2010) for the same cultivar.
The mean values of seed volume of common bean cultivars determined by the replacement method were given in Table 2.The seed volume value was the highest (0.531 cm 3 ) in 'Çayırlı' and the lowest (0.137 cm 3 ) in 'Karacaşehir-90'.The second highest seed volume value (0.336 cm 3 ) was determined in 'Elkoca-05'.'Göynük-98' and 'Yakutiye-98' had the almost close values of 0.345 cm 3 and 0.346 cm 3 , respectively.Fıratlıgil-

225
Durmuş et al (2010) determined the seed volume using the volume equation of ellipsoid, compared these calculated values to the experimental volume values, and reported that the values were close to each other.In our study, the difference between replacement method and calculation method was the highest (7.2%) for 'Çayırlı', which had the highest seed mass value and the lowest (0.7%) for 'Karacaşehir-90', which had the lowest seed mass value.
The means of projected area, equivalent diameter and perimeter determined at the width orientation were higher than those of the thickness orientation (Figure 4).In this study, the minimum and maximum values of projected area, equivalent diameter, and perimeter were 32.3-222.9mm 2 , 6.4-16.8mm, and 22.3-61.6mm for width orientation, and 26.4-130.2mm 2 , 5.8-12.9mm, and 20.6-54.3mm for thickness orientation.These data are in agreement with those reported by Fıratlıgil-Durmuş et al (2010) for seven common bean cultivars (projected area: 30.5-167.2 mm 2 , equivalent diameter: 6.2-14.6 mm, and perimeter: 20.5-99.4mm).The value of projected area ranged between 56 and 116 mm 2 at the width orientation for 91% of all seeds and 26 and 86 mm 2 at the thickness orientation for 96% of all seeds.The value of equivalent diameter ranged between 6.8 and 12.2 mm at the width orientation for 91% of all seeds and between 6.2 and 10.4 mm at the thickness orientation for 94% of all seeds.The value of perimeter ranged between 24.2 and 45.2 mm at the width orientation for 91% of all seeds and between 22.1 and 43.1 mm at the thickness orientation for 96% of all seeds.
The projected area, equivalent diameter, and perimeter values determined at the width and thickness orientations were presented in Table 3.The highest values of projected area, equivalent diameter, and perimeter were obtained with 'Çayırlı' while the lowest values with 'Karacaşehir-90' at both orientations.The data showed that the relationship between the means of the gravimetric and the geometric properties of the cultivars was a positive association.The differences between the means
The mean values of length, width, and thickness of common bean cultivars were given in Table 4.The minimum and maximum values ranged between 7.8 and 22.6 mm for length, 4.9 and 11.8 mm for width, and 2.3 and 7.7 mm for thickness.The highest value of length was obtained in 'Çayırlı' (17.8 mm), followed by 'Elkoca-05' (15.3 mm).The mean values of length were similar to each other for 'Göynük-98' (14.6 mm), 'Yakutiye-98' (14.4 mm), and 'Yunus-90' (14.6 mm).In addition, 'Önceler-98' (11.8 mm) and 'Terzibaba' (11.7 mm) had a close value.Güvenç & Güngör (1996) reported that the mean values of length for seven common bean cultivars varied between 8.6 and 15.0 mm, supporting our results.In the literature, 'Barbunia' followed the 'Çayırlı' with a length of 16.68 mm (Cetin 2007).'Karacaşehir-90' had the lowest mean value in terms of length (9.1 mm) and width (5.8 mm).'  had the same mean value of width (7.6 mm).Moreover, the value of width of '  mm.The highest value of width (10.0 mm) was measured in 'Çayırlı'.The highest value of thickness was obtained with 'Elkoca-05' (6.0 mm) and 'Noyanbey-98' (6.0 mm) as the lowest value was obtained with 'Terzibaba' (4.6 mm).'  had the same value of thickness value of 5.8 mm.Cetin (2007) reported the thickness value of 'Barbunia' as 7.51 mm at the moisture content of 18.33%.As seen from Table 4, the common been cultivars were classified according to their size and shape (Şehrali 1988).Six cultivars were oblongus, two cultivars were ellipticus, three cultivars were subcompressus, and 'Çayırlı' genotype was compressus.
As physical properties of agricultural materials, the geometric mean diameter and surface area are functions of the values of length, width, and thickness.Sphericity, defined as the ratio of geometric mean diameter to the greater diameter
When the shape factor is one, it means that the seed is in circular form in a plane.In the study, the shape factor values of seeds were determined at two orientations, i.e., width and thickness.As can be observed in Table 5, the mean values of shape factor of the cultivars obtained at the thickness orientation were in agreement with the mean values of sphericity.
Figure 5 shows the mean elongation values of the common bean cultivars calculated for width, thickness, and vertical orientations.'Terzibaba', shape factor of which is the highest value at the width orientation, had the lowest elongation value at the same orientation.While the shape factor of 'Elkoca-05' was the lowest mean value at the width orientation, its elongation value had the highest mean value at the same orientation.These results showed that there was a negative relationship between the shape factor determined at the width orientation and elongation values.The lowest elongation value (1.86) was obtained from thickness orientation for 'Karacaşehir' cultivar.The elongation value obtained from thickness orientation was the highest (3.42) in 'Çayırlı' cultivar.The means of shape factor of both cultivars determined at the thickness orientation was inversely correlated In order to explain the relationship between variables and detect the most important factors of variability, principal component analysis was applied to mean values.The principal components (PCs) showing the proportion and cumulative variance of the variables are presented in Table 6.The first three components, namely PC1, PC2, and PC3, formed 76.3%, 19.8%, and 3.2% of the variation, respectively, for a total of 99.3%.Considering the size properties, it can be concluded that the values of length, width and thickness are correlated with the components PC1, PC2 and PC3.The variables of length and width had positive correlation with their components as the thickness variable had a negative correlation with its component.
The most important variables of PC1 are seed mass, geometric mean diameter, surface area, volume, projected area at two orientations, equivalent diameter, perimeter and length.PC1 was positively correlated with all variables.It can be said that these variables can ensure the grouping common bean cultivars in terms of the geometric and gravimetric properties.On the other hand, PC1 includes only length variables among the size variables and this means that the length variable is more important in grouping the cultivar than the other size variables.The elongation values determined at the length orientation, shape factor and width values determined at the width orientation are the most important variables of PC2, having a positive correlation.The elongation values determined at the width orientation is important variable of PC2  In order to explain the relationship between variables and detect the most important factors of variability, principal component analysis was applied to mean values.The principal components (PCs) showing the proportion and cumulative variance of the variables are presented in Table 6.The first three components, namely PC1, PC2, and PC3, formed 76.3%, 19.8%, and 3.2% of the variation, respectively, for a total of 99.3%.Considering the size properties, it can be concluded that the values of length, width and thickness are correlated with the components PC1, PC2 and PC3.The variables of length and width had positive correlation with their components as the thickness variable had a negative correlation with its component.
The most important variables of PC1 are seed mass, geometric mean diameter, surface area, volume, projected area at two orientations, equivalent diameter, perimeter and length.PC1 was positively correlated with all variables.It can be said that these variables can ensure the grouping common bean cultivars in terms of the geometric and gravimetric properties.On the other hand, PC1 includes only length variables among the size variables and this means that the length variable is more important in grouping the cultivar than the other size variables.The elongation values determined at the length orientation, shape factor and width values determined at the width orientation are the most important variables of PC2, having a positive correlation.The elongation values determined at the width orientation is important variable of PC2 having a negative correlation.The variables in PC2 can ensure grouping the cultivars depending on shape factor and width values.In PC3, the shape factor values determined at the thickness orientation had a positive correlation.

231
having a negative correlation.The variables in PC2 can ensure grouping the cultivars depending on shape factor and width values.In PC3, the shape factor values determined at the thickness orientation had a positive correlation.
Figure 6 shows the Eigen scores graph related to PC1 and PC2 explaining an important portion of variation.As can be observed in graph, 'Çayırlı' was significantly different from the other cultivars in terms of size, shape, and gravimetric properties.It can be said that 'Aras-98', 'Yakutiye-98', 'Göynük-98', 'Noyanbey-98', 'Yunus-90', and 'Elkoca-05' took place in a second group in respect with their geometric and gravimetric properties.As a third group, 'Terzibaba', 'Kantar-05', and 'Akman-98' had lower length value, higher width value, and lower thickness value than the other cultivars.The cultivars of the third group had also more circular shape at the width orientation.In the   4. Therefore, the estimates of PCA support the results of cultivar classification according to Şehrali (1988).

Conclusions
Based on the results of this study, the following conclusions could be drawn: • 'Çayırlı' genotype had the highest mean value in terms of some geometric and gravimetric properties such as seed mass, volume, geometric mean diameter, surface area, projected area, equivalent diameter, perimeter, length and width while 'Karacaşehir-90' had the lowest value.
• The thousand seed mass value of 'Çayırlı' was 3.65 times higher than that of 'Karacaşehir-90'.
• The mean values of sphericity of the cultivars investigated varied in the range of 58.7% -70.0%.'Karacaşehir-90' had the highest value of sphericity.
• 'Terzibaba' was closer to the circular shape at the width orientation.Conversely, 'Elkoca-05' had the lowest value of shape factor.At the thickness orientation, the highest and lowest values of shape factor were measured in 'Karacaşehir-90' and 'Çayırlı', respectively.
• The results of principal component analysis were in agreement with those of estimates of variance

Figure 2 -Figure 3 -
Figure 1-Bean cultivars Şekil 1-Fasulye çeşitleri Ta r ı m B i l i m l e r i D e r g i s i -J o u r n a l o f A g r i c u l t u r a l S c i e n c e s 19 (2013) 219-234 224 obtained from the mean values of the size and shape properties of the cultivars and the output data sets including Eigen values and vectors and principal component scores.
Türkiye'de Tescilli Fasulye (Phaseolus vulgaris L.) Çeşitlerinin Dijital Fotoğraflama Kullanılarak Boyut ve Şekil Analizi, Kara et al Ta r ı m B i l i m l e r i D e r g i s i -J o u r n a l o f A g r i c u l t u r a l S c i e n c e s 19 (2013) 219-234

Figure 4 - 234 227
Figure 4-Numerical distribution of some geometric properties determined at two different orientations of seed Şekil 4-Tohumun iki farklı oryantasyonunda belirlenen bazı geometrik özelliklerin sayısal dağılımı Türkiye'de Tescilli Fasulye (Phaseolus vulgaris L.) Çeşitlerinin Dijital Fotoğraflama Kullanılarak Boyut ve Şekil Analizi, Kara et al Ta r ı m B i l i m l e r i D e r g i s i -J o u r n a l o f A g r i c u l t u r a l S c i e n c e s 19 (2013) 219-234

Seed
Size and Shape Analysis of Registered Common Bean (Phaseolus vulgaris L.) Cultivars in Turkey Using Digital Photography, Kara et al Ta r ı m B i l i m l e r i D e r g i s i -J o u r n a l o f A g r i c u l t u r a l S c i e n c e s 19 (2013) 219-234 232 fourth group, 'Önceler-98' was closer to the axis of PC1 because it had higher values of seed mass and volume than 'Karacaşehir-90'.'Karacaşehir-90' took place at the farthest point in the left side of PC1 axis, due to its lowest mean values of variables in PC1 component.These cultivar groups determined by PCA (Figure 6) were in agreement with the classification presented in Table

Table 1 -Moisture content and some gravimetric properties of bean cultivars (mean±standard deviation)
Çizelge 1-Fasulye çeşitlerinin bazı kütlesel özellikleri ve nem içeriği(ort.±standartsapma) * Means followed by the same letter in the column are not different as determined by the Duncan test at a 5% significance level.

Table 2 -Mean values of seed volume determined by the replacement method and calculation method and the differences between two methods (mean±standard deviation)
* Means followed by the same letter in the column are not different as determined by the Duncan test at a 5% significance level.

Table 4 -Length, width and thickness values of the bean cultivars (mean±standard deviation) and classification regarding the shape properties
* Means followed by the same letter in the column are not different as determined by the Duncan test at a 5% significance level

Table 5 -Means of geometric mean diameter, surface area, sphericity and shape factor of bean cultivars (mean±standard deviation)
Means followed by the same letter in the column are not different as determined by the Duncan test at a 5% significance level Seed Size and Shape Analysis of Registered Common Bean (Phaseolus vulgaris L.) Cultivars in Turkey Using Digital Photography, Kara et al * 11the same orientation, the elongation values of 'Elkoca-05', 'Aras', 'Terzibaba', 'Göynük',  'Noyanbey',' cultivars ranged from 2.47 to 2.60, values of which were considerably close each other.Shape factor obtained from thickness orientation for the same cultivars ranged from 0.65 to 0.67.The elongation values obtained from the vertical orientation changed 1.19 to 1.92.'Karacaşehir-90' had the lowest elongation value while the highest elongation value was determined at the vertical orientation for 'Çayırlı' cultivar.