Methods of Measuring Bruise Volume of Pear (Pyrus Communis L.)

Abstract

The grade of injures can be determined or measured by bruise volume using a sectioning and imaging analysis technique and different methods for calculating the size of pear bruises. The calculation methods were not consistent, with large calculation errors at small bruise sizes. The calculation of bruise volume was improved when the surface shape of the bruise was an ellipse instead of a circle. In this study, various methods are recommended for the calculation of the bruise volume depending on the range of the bruise sizes being investigated.

Keywords:

• Bruise
• Volume
• Pear
• Compare

INTRODUCTION

Fruit bruising is one of the most important factors limiting the mechanization and automation in harvesting, sorting and transport of soft fruits and vegetable, Dark spots appearing near the product surface are due to the previous mechanical contacts of the products with other bodies. Force loading the fruit can be very variable ranging from static to dynamic. Bruise extent is usually described in terms of bruise volume, which is in a tight relation to the product quality.

In modern agriculture and competitive markets, the importance of fruit quality is significant. More than 20% of pears are injured during common practice of picking, sorting, grading, packing, transportation and storage. These injuries cause quality reduction and loss of profit. The damage appears as fruit injuries such as light bruises, scratches, cuts and pits. In an effort to minimize the bruise, a study was conducted to investigate sources of mechanical injuries causing fruit quality loss. The goal of the research was to explore at what stage and to what level of severity, mechanical injuries occur during fruit handling from the tree to the commercial sorting.

It was estimated that 30% to 40% of fruits and vegetables undergo mechanical damage from harvesting to market delivery.[1] Fruit and vegetables are highly susceptible to damage during harvesting, handing, transportation and storage. The damage would make them spoil quickly, reduces quality and increased the loss.[2] The most common symptom of damage is fruit bruising. Dark spots appear near the product surface are due to previous forceful mechanical contacts of the products with other bodies. Bruise extent is usually described in terms of bruise volume,[3] which closely relates to product quality. Studman[4] lists fourteen factors affecting bruising of apples, but the role of some of them is slightly controversial.

Bruising is initiated by the breakage of cell membranes[5] leading cytoplasmic enzymes to react with vacuolar contents.[6] Partington et al.[7] reported that bruised tissue undergoes cell death in contrast to wounded tissues showing continued induction of defense proteins, tissue repair and no cell death. About 3 h after impact, cell death was initiated and coincided with de-compartmentalization, the appearance of lipid peroxides and melanin production. Finally, bruising induces death and dehydration of all the cells, which leave a large space indicating loss of effective barrier against infection.[7]

Bruise extent is usually described in terms of bruise volume, which effects product quality.[3] To determine the damage susceptibility of pears, research is needed using the size of the bruises. It is difficult and time consuming to measure bruise volume and to create accurate data. To decrease this time, it is necessary to make some assumptions about the bruise shape and take measurements only of the dimensions.[8–10] Several different models have been proposed to describe the bruise shape. In this study, the optimal method was determined for calculating the bruise size for pears.

MATERIALS AND METHODS

Ankara pears (Pyrus communis L.) were used for all the experiments in this study. The fruits were collected from local garden in the Turkey during the autumn season of 2005. Harvested fruits were transferred to the laboratory in polythene bags to reduce water loss during transport. The fruits were cleaned in an air screen cleaner to remove all foreign matters such as dust, dirt and chaff as well as immature and damaged fruits. The initial moisture content of fruits was determined by using drying oven method.[11] The remaining material was packed in a hermetic vessel and kept in cold storage at 0°C until used for experiments.

Uniform size Ankara pears (mean weight 150 ±14 g) were removed from 4 weeks of storage and impacted using a pendulum impacting device based on the method discussed by Mohsenin.[9] Every pear tested dynamically by an impact pendulum. The pendulum has a 60 cm long arm with removable weight and changeable impactors with a flat and a spherical head of diameter 15 mm. The pendulum arm was then fixed in one of the initial positions and dropped on the fruit. After rebounding of the arm into the highest position, the arm was caught by hand. Ten fruits were dropped at five heights of 25, 50, 75, 100, and 125 mm, respectively.

CALCULATION OF BRUISE VOLUME METHOD

Most research on pear damage use impact energies which result in all the flesh between the skin and the lower bruised boundary being damaged. The shape is usually assumed to have a spherical boundary. However, in studies using more realistic energies that occur in practical handling systems.[12 –13] The bruise very rarely extends to the fruit surface, and can often have a poorly defined spherical shape. It was therefore unclear how well these volume calculation methods could predict bruise volume for bruises more typical of modern handling systems.

The methods used are full depth method, bruise thickness method, ellipsoid method, unbruised volume removed method and enclosed volume method. These different methods which have been proposed to describe bruise volume are shown in Table 1. The formulas have been retained in a form recognizable from the original authors, but common nomenclature has been used below. The measurements are the depth, d_b and width, w₁, across the major axis of the bruise (Fig. 1). Other measurements which can be required by some methods are the bruise width across the minor axis, w₂ and the depth to the top of the bruise, d_t (Fig. 1).

Table 1 Formula of five methods for calculation of bruise volume

		Formula of bruised volume
Method of bruised volume calculation	Figure of bruised volume	Circle	Ellipse
Bruise thickness
Full depth
Enclosed volume
Unbruised volume removed
Ellipsoid

Figure 1 Measurements taken to calculate bruise size.

Full-depth Method

The bruise thickness method requires a measurement to be made of the bruise thickness at the contact plane, but as the fruit recovers its shape somewhat upon rebound, it is not clear exactly where this measurement is made. It is likely that in most cases the depth/thickness measurement which is made as the full depth from the surface.[14] The assumed shape, therefore, is a section of a sphere based on the full depth as shown in Table 1.

Bruise Thickness Method

A section through the bruise at its midpoint, perpendicular to the skin, appeared to have a circular profile. The bruise volume could be described as a section of a sphere.[9] The volume was calculated for a spherical shape defined below the contact plane (point of maximum defection of the pear during impact) (Table 1). For bruising at low impact energies a simple approximation for the estimate of the position of the contact plane is to assume that this occurs at the same depth as the top of the bruised tissue. Observation of bruise shapes suggests that this is a reasonable approximation. The section of a sphere can thus also be calculated, based on the bruise thickness (Table 1).

Ellipsoid Method

Another style, which has been used to describe other bruising of fruit, is to approximate the bruise as an elliptical shape,[10] and this method was also included in this evaluation (Table 1).

Unbruised Volume Removed Method

A modification to the Holt and Schoorl model has also been used[13–15] which assumes a spherical shape of undamaged section above the bruise (V₂+V₃) as shown in Table 1. This volume is subtracted from the enclosed volume calculation.

Enclosed Volume Method

Holt and Schoorl[8] purposed a modification to the spherical shape model where they suggested the volume be estimated as the sum of the two volumes; the volume below the contact plane, bounded by the bruise boundary, V₁, and the volume bounded by the fruit surface, V₂ (Table 1). The calculation of the second volume required an additional measurement to be made to determine the radius of curvature of the fruit surface at the bruise site.

Alternative Bruise Shape

The original theory for all the above volume estimation methods assumed a circular bruise shape when viewed perpendicular to the fruit surface. Some of the pear bruising observed in this study had a distinct elliptical shape, dependent on the surface shape of the Ankara pears. The volume estimate formula were therefore, also, all modified to provide an elliptical estimate, based on the major and minor axes, as well as the circular assumption and these are also shown in Table 1. The radius of curvature (Fig. 2) at each impact site was determined by the following equation.[9]

(1)

Figure 2 Determining radius of curvature with equation.

The fruit was keep for 24 h at room temperature to allow the browning color to develop. The fruit was then separated at each bruise site. First, a section containing the bruise and surrounding tissue was taken from the fruit. This section was then sliced into 1 mm thick parallel slices. These were taken so that the plane of the central slice extended into the fruit perpendicular to the skin surface at the centre of the bruise and parallel to the stem-calyx axis. The slices for each bruise were then placed on a prepared surface alongside a 1 cm² square template. An image was captured using a digital camera (Kodak DC5000).

The image was analyzed using Adobe PhotoShop 6.0 and Global Lab Image. The extremity of the 1 cm² square template was traced and the area calculated in pixels, to provide a pixel/mm² calibration for each image. The average resolution was 840 pixels/mm². The browned area on each slice was then manually traced, using the selection tool in the software. The areas for each slice were calculated and then summed, to determine a total bruised area for the slices. For the range of impacts used, the total areas ranged from 4150 to 223 364 pixels. The volume was then calculated by multiplying by the slice thickness. In addition, the width across the major axis w ₁, depth from the skin to the top d_t and the bottom d_b of the bruise were measured for the central slice of the bruise, to conform to the measurements made by the other five bruise evaluation techniques. The width of the bruise across the minor axis w ₂, perpendicular to the stem-calyx axis, was calculated as the product of the number of slices and the thickness of the slices.

Methods of bruise volume calculation were shown in above Table 1 where, V is the calculated bruise volume (mm³), d_b is the full depth of bruise (mm), d_t is the depth from fruit surface to top of bruise (mm), R is the fruit radius at bruise (mm) (Fig. 1), V₁, V₂, and V₃ are the volumes used in calculation of V (mm³), w ₁ and w ₂ are the bruise widths across the major and minor axes) (mm) (Fig. 1). The values of x and y dimensions of bruise were calculated by using the following relationship:

(2)

(3)

where, x is the height of fruit section above the contact plane (circular bruise) (mm), y is height of fruit section above the contact plane (elliptical bruise) (mm), R is the fruit radius at bruise (mm); and w ₁ and w ₂ are the bruise widths across the major and minor axes (mm).

RESULTS AND DISCUSSION

The different bruise calculation methods produce a considerable range of estimated volumes when compared with the measured volume, with some overestimating by up to 90% for certain bruise sizes. For bruises above 400 mm³, the different calculation methods differed by approximately 500 mm³. When compared to the measured volume, a good volume calculation method needs to have a slope of unity and an intercept of zero. The methods were also compared to determine the merits of the assumptions that bruises were circular or elliptical. In addition, the predictions were compared for two sets of the bruise data all impacts and a subset of low impacts. Table 2 shown that the summary of regression fits for the different methods. When comparing regression slopes the bruise predictions fall into two groups which were consistently and significantly different (p < 0.05); the ellipsoidal, full depth and enclosed volume methods produce similar slopes which are steeper than the bruise thickness and unbruised volume removed methods (Table 2). In all bruise calculation methods expressed by the coefficient of determination R ² changed from between 0.90 and 0.98.

Table 2 Linear regression for volume calculation methods

	All impacts						Low impacts
	Circular			Elliptical			Circular			Elliptical
Method	m	C	R^2	m	C	R^2	m	C	R^2	m	C	R^2
A	0.98	50.9	0.94	0.98	−0.87	0.98	1.50	−38.9	0.90	0.87	−32.1	0.91
B	1.32	123.1	0.93	1.22	46.2	0.97	1.84	−11.5	0.95	1.35	1.9	0.94
C	1.23	141.8	0.93	1.26	63.4	0.97	1.74	−8.1	0.90	1.42	6.9	0.98
D	0.97	82.1	094	0.93	14.7	0.98	1.50	−31.9	0.95	0.91	−21.2	0.95
E	1.32	79.8	0.94	1.23	−2.4	0.98	1.76	−43.2	0.91	1.15	−31.8	0.95
A: bruise thickness; B: full depth; C: enclosed volume; D: unbruised volume removed; E: ellipsoid; m: slope of regression line; C: constant; and R^2: coefficient of determination.

Based on the impacts for all the five drop heights and assuming a circular bruise calculated using the bruise width parallel to the stem-calyx axis w₁ , three methods overestimated the bruise volume. The unbruised volume removed and bruise thickness methods provide slopes, of 0.95 and 0.96. These slopes were not significantly different from a slope of 1 (p < 0.05). The bruise thickness method produced an intercept closer to zero. When both widths are used in the calculation, assuming an elliptical shape of bruise, the best slope estimate was the enclosed volume, which was not significantly different from a slope of 1 (p < 0.05). The unbruised volume removed and the bruise thickness methods both provided reasonable estimates but with a slope approximately 15% low. The ellipsoidal and full-depth methods over predict slope by approximately 15%. The ellipsoidal and bruise thickness methods both provide intercepts near to zero (Table 2). When the bruise shape was assumed to be circular, the ellipsoidal, full depth and enclosed volume methods over predict bruise size with slopes of 1.97–2.24. The bruise thickness and unbruised volume removed methods also over-predict bruise volumes, both with slopes of 1.55. All five slopes were significantly greater than 1 (p < 0.05). When the shape of the bruise is assumed to be elliptical, the bruise thickness and unbruised volume removed methods had slopes which were not significantly different from 1 (p < 0.05), but with negative intercepts. This implies that they consistently under predict bruise volume by 9–18 mm³. The slopes for the other three methods mean that they progressively over predict bruise volume by approximately 22% (Table 2).

CONCLUSIONS

All the volume estimation methods induce errors in prediction of actual bruise size. There was not one single method which can be used for the estimation of bruise volume over the range of commercially significant impacts. Treating the bruises as an elliptical shape improved the accuracy (linearity) of the volume estimates. The enclosed volume and full-depth methods assume the bruise extends to the surface. In addition, the full depth method includes a volume, which is outside the fruit. This was reflected in the assessment as it consistently predicted one of the steepest slopes and highest intercepts. The full-depth method is not an appropriate method for estimating bruise volume. The ellipsoidal method calculated the volume based on the actual bruise depth, but over predicted volumes to a similar level to the full-depth method. This implies that the assumption of a basically spherical-shaped bruise, even in cases where the actual bruise shape is poorly defined, is still a valid assumption when calculating the bruise volume. The assumption made for the bruise thickness method, that the contact plane was at approximately the same height as the top of the bruise provided a method with very similar estimates to the unbruised volume removed method. The slopes were virtually identical and the difference in the intercepts was on the order of only 15 mm³. Both these methods calculate a volume which is closer in profile to the actual observed bruise, and could therefore be expected to provide a better prediction. The simpler technique is the bruise thickness method which does not require the additional measurement of fruit radius. So, when considering an appropriate technique for estimating pear bruising, where bruise widths can be measured across both the major and minor axes, the enclosed volume method provides the best estimation of volume for the full range of impacts (less than 100 mm drop). For the small impacts (less than 50 mm drop), the bruise thickness method provides the best estimates of volume. For bruising evaluations, it is often difficult to measure both widths as well as depth in practice and so, based on bruise sections parallel to the stem-calyx axis, the bruise thickness method provides the best estimation method for bruising which occurs over the full range of impacts. For studies focused solely on small impact energies (50 mm drops and lower), it may be appropriate to use the bruise thickness method and scale volumes to obtain a more accurate estimate, where only one bruise width is measured as all methods over predict the actual volumes.

ACKNOWLEDGMENTS

This study was partly supported by the Scientific Research Fund of Akdeniz University, Antalya, Turkey.