JP5348798B2 - Relative evaluation method based on pattern recognition of food hardness, texture and texture - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To exactly grasp difference in hardness, texture, and attribute of food in mastication. <P>SOLUTION: A sample specimen (A) of the food is pressed by a plunger, weight and distortion rate are continuously measured, a distortion rate-weight curve of a multidimensional approximation curve such as five or more dimensional approximation curve which considers an X axis as the distortion rate, and a Y axis as the weight by a least square method based on values of the weight and the distortion rate, and the maximum value (MaxA) in the multidimensional approximation curve is calculated. A distortion rate-weight curve of the multidimensional approximation curve of the same dimension is created for a test specimen (B) of the food, and the maximum value (MaxB) in the multidimensional approximation curve is calculated. After that, a corrected multidimensional approximation curve in which weight values on the multidimensional approximation curve is integrally corrected is created for each of the multidimensional approximation curves so that the weight values of the maximum value (MaxA) and the maximum value (MaxB) become the same value, a difference between the two created corrected multidimensional approximation curves is integrated, and a distinction in the hardness, texture and attribute between the sample specimen (A) of the food and the sample specimen (B) of the food is evaluated from the integrated value. <P>COPYRIGHT: (C)2012,JPO&amp;INPIT

Description

  The present invention relates to a relative evaluation method based on pattern recognition of food hardness, texture, and texture. Specifically, for a variety of products such as agar, gelatin, tofu, cookies, biscuits, and pies, pattern recognition is performed by numerically quantifying differences in hardness, texture, and texture perceived when chewing food with a small number of measurements. It is related with the method for evaluating relatively.

  For human beings, food intake is not just an act of acquiring energy to sustain life, but more positively by using all five senses of taste, smell, touch, sight, and hearing. Is an act that seeks to enjoy "satisfaction" and "happiness". In general, “taste”, “fragrance” and “appearance” as well as “texture” are listed as elements of “taste” of food.

Texture, as defined by the International 0rganization for Standardization, is the sum of the rheological and structural attributes of food that can be perceived mechanically, tactilely and, if appropriate, visually and audibly. It is supposed to be. That is, texture is generally understood as a term that means both the texture that is sensed and expressed by a person who eats food and the physical properties of the food itself. For texture measurement or evaluation, conventionally, sensory evaluation for evaluation by human sensory organs and physical measurement for objective evaluation of physical properties of foods have been performed.
To date, however, there has been no established and widely accepted method for measuring or evaluating the overall texture of foods.

  In general, as a measuring device for physical properties of various foods, for example, a device called a rheometer, a creep meter, or the like that measures mechanical properties is widespread. The apparatus can measure compressive breaking strength, tensile strength, cutting strength, elasticity, viscoelasticity, brittleness, adhesiveness, stress relaxation, creep, and the like.

  Several proposals have been made regarding methods for measuring physical properties of foods or methods for evaluating texture. For example, for foods for infants or those with difficulty in swallowing (mousse), based on the measurement of the shape of the upper jaw model, using a food property measuring instrument equipped with a palatal container and tongue plunger that modularize the shape of the palate and tongue, respectively, There is a proposal (Patent Document 1) for measuring stress.

  In addition, use a rheometer to obtain the breaking curve of test foods such as cookies and snacks, obtain the breaking energy in a predetermined frequency range by mathematical analysis, and perform statistical analysis with the crispness of the sensory test. Therefore, there is a proposal (Patent Document 2) as an index of crispness.

  In addition, gel-forming foods such as tofu, koji, and cheese whose physical properties change from sol to gel are irradiated with light having a specific wavelength (ranging from 400 nm to 50,000 nm) and the resulting absorbance curve is obtained. In particular, there is a proposal for a method for judging the quality of gel-formed foods utilizing the fact that there is a high negative correlation between the absorbance at 800 nm to 840 nm and the breaking force (Patent Document 3).

  Also, a rheometer probe is inserted into food such as kiwi or celery, the vibrations generated are obtained, and the vibration data from which noise has been removed is obtained as the amplitude density per unit time. The higher the amplitude density, the more "crispy" There is a proposal (Patent Document 4) that evaluates that the leeks are higher than the Japanese radish (the leeks have higher amplitude density and are more crisp).

  In addition, the layered foods such as croissants and Danish pastries are suppressed with a rheometer, and the total load applied to the plunger is calculated as a breaking energy value E to obtain “breaking energy value E / number of breaking points N”. There exists a proposal (patent document 5) evaluated as an index of the texture of layered food.

  Also, load and strain rate data for multiple samples of food, for example gel foods such as agar, gelatin, nata de coco, konjac gel, fruits such as apples, pears and bananas, baked confectionery such as cookies, biscuits and pies Is measured using a device (rheometer, etc.) that can continuously measure the quaternary equation approximate curve or the quaternary equation approximate curve based on the obtained load and strain rate measurements. Proposed by the method of least squares and evaluating the inclination of the tangent at the inflection point of the curve portion before reaching the local maximum value that is the breaking point in the curve as the hardness of the texture (Patent Document 6, Patent Document 7) )

JP 2000-283975 A JP 2001-133374 A JP 2003-106995 A JP 2007-57476 A JP 2007-225460 A Japanese Patent Application No. 2009-70951 Japanese Patent Application No. 2010-85178

  However, the proposal of Patent Document 1 is merely an improvement of the method for measuring the maximum stress by modularizing the shape of the human palate and tongue. In addition, the proposal of Patent Document 2 is difficult to apply in determining the difference between cookies and snacks having similar sound patterns while having different textures. Moreover, the proposal of patent document 3 only confirms the physical property of the foodstuff which changes from sol to gel, and does not evaluate the food texture of the obtained baked confectionery. In addition, the proposal of Patent Document 4 can evaluate the “crispness” of vegetable tissues such as leeks and radish, but determines the difference in texture of cookies and snacks having similar texture patterns while having different textures. Is difficult to apply. In addition, the proposal of Patent Document 5 is a proposal for evaluating the texture of layered foods such as croissants and Danish pastries, and is effective for layered foods such as pie in baked confectionery, but is suitable for cookies, snacks and the like. Considering the case of application, it was not effective for all baked goods. In addition, the proposal of Patent Document 6 is effective for a gel-like food having a relatively simple breaking behavior, but in the case of baked confectionery, the breaking occurs rapidly, and the behavior of the strain rate-load curve is complicated. In some cases, the measured values of load and strain rate deviate from the approximate quaternary equation, and the approximation of the quaternary equation was not appropriate and effective.

  In addition, as a problem common to the techniques described in Patent Documents 1 to 5, there is a point that a method of patterning and relatively evaluating differences in food hardness, texture, and texture has not been established. . Furthermore, with respect to the technique described in Patent Document 6, even if an attempt is made to evaluate the relative differences by digitizing differences in food hardness, texture, and texture by recognizing patterns from the obtained approximate quaternary equation, load In addition, since the measured value of the distortion rate may deviate from the approximate quaternary equation, the evaluation is difficult.

Therefore, the technique described in Patent Document 7 has been improved by making the approximate quartic equation described in Patent Document 6 an equation of the fifth or higher order. As a result, according to the technique of Patent Document 7, by preparing a fifth-order or higher order approximate curve with a small degree of deviation from the actually measured load and strain rate measurement values, gel food, baked goods It is possible to obtain an index that reflects the hardness, texture, and texture actually received over a wide variety of foods such as foods.
However, it is left to the judgment of the measurer to compare and evaluate the relative differences in hardness, texture, and texture among multiple foods by comparing approximate curves of fifth-order or higher-order polynomial equations. The accuracy and reliability of the evaluation was lacking.

Under such circumstances, it has been desired to develop a new evaluation method that can objectively compare differences in hardness, texture, and texture among a plurality of foods.
Therefore, the present invention relates to various foods such as baked confectionery and gel-like foods by using an index that accurately reflects the difference in the hardness, texture and texture of foods that humans actually sense when chewing foods. The degree of hardness, texture and texture of food during chewing are quantified numerically, and the obtained quantitative values are used to relatively evaluate the differences. And a relative evaluation method based on texture pattern recognition.

According to the prior art, in baked confectionery such as cookies, biscuits and pies, gelatinous foods such as gelatin, agar, tofu, and many other foods, the maximum value, minimum value, and inflection of the strain-load curve It is considered that the inclination of the point or the like serves as an indicator of the hardness, texture and texture of the food, and corresponds to the hardness, texture and texture quality sensed when a human actually chews the food. That is, it is considered that the shape of the strain rate-one load curve expresses the hardness, texture, and texture quality.

Therefore, the present inventor has determined the results of strain rate one load measurement as a basis for obtaining an index of hardness, texture, and texture quality when chewing baked confectionery, gel food, or other foods. Was approximated with a multi-order equation of the fifth or higher order, which is one of the most effective approximations, and a multi-order equation approximate curve of a strain rate-one load curve was created. Moreover, in order to compare the quality of hardness, food texture, and texture, the strain rate-one-load fracture was measured for a plurality of samples, and a multi-order equation approximate curve of the strain-rate load curve was created. And from this multi-order equation approximation curve, the difference of hardness, texture and texture between multiple samples is automatically calculated, and the obtained calculated value is the difference in quality of hardness, texture and texture. As a result, we have intensively studied an analysis method that can discriminate differences in hardness, texture, and texture by automatically recognizing patterns.

As a result, for each of the above approximated multi-order equation approximate curves, a corrected multi-order approximate curve in which the load values on the approximate curve are uniformly corrected is created, and the difference between the obtained corrected multi-order approximate curves is integrated and calculated. Thus, the present inventor has found that the above-mentioned problem can be solved by adopting a method of quantifying the degree of divergence between the approximated polynomial equations.

In addition, the above-mentioned method can express, for example, how much the texture deteriorates immediately after opening for baked confectionery, so in the same food, the change in hardness, texture, and texture with time The present inventor has found that the degree can be grasped numerically and applied to quality control of food.

Thus, in the present invention, the food sample (A) is pressed with the plunger, and simultaneously, the load and the strain rate during pressing are continuously measured. Based on the values of the load and the strain rate, the least square method is used. Perform a calculation to create a strain rate-one-load curve of a fifth order or higher order approximate curve with the X axis as the strain rate and the Y axis as the load, and calculate the maximum value (MaxA) in the multiorder approximate curve. Seeking
On the other hand, for the food sample (B), in the same manner as described above, a strain rate-one load curve of a homogeneous approximation curve is created, and the maximum value (MaxB) in the approximation curve is calculated and obtained.
Next, a corrected multi-order approximate curve obtained by uniformly correcting the load values on the multi-order approximate curve so that the load values of the maximum value (Max A) and the maximum value (Max B) are the same value is obtained as the multi-order approximate curve. Create for each of the fitted curves,
Next, the difference between the two corrected multi-order approximate curves prepared is integrated, and the obtained integrated value is used to determine the hardness, texture, and texture of the food sample (A) and the food sample (B). The present invention provides a relative evaluation method based on pattern recognition of food hardness, texture, and texture, characterized by evaluating differences.

  According to the present invention, with regard to baked confectionery such as cookies, biscuits and pies, gelatinous foods such as gelatin, agar, and tofu, or other foods, the hardness, texture and texture of the food when chewing the food It is possible to accurately grasp the difference in quality and statistically distinguish it with high accuracy.

It is a figure which shows an example of the distortion-load curve of the multi-order approximation curve created based on the data of the load and distortion of a food sample. It is a figure which shows the example of an area | region of the difference of a multi-order approximate curve. It is a figure which shows the 5th order approximate curve (A) and corrected 5th order approximate curve (B) of cookie A. Fifth order approximation curve cookies A ₂ (A), is a diagram illustrating a correction fifth order approximation curve (B). It is a figure which shows the 5th order approximate curve (A) of biscuit B, and a correction | amendment 5th order approximate curve (B). It is a figure which shows the 5th order approximate curve (A) of biscuit C, and a correction | amendment 5th order approximate curve (B). It is a figure which shows the 6th approximation curve (A) and corrected 6th approximation curve (B) of cookie A. Next six trendline Cookie A ₂ (A), is a diagram illustrating a correction six linear approximation curve (B). It is a figure which shows the 6th approximation curve (A) of biscuit B, and a correction | amendment 6th approximation curve (B). It is a figure which shows the 6th order approximate curve (A) of biscuit C, and a correction | amendment 6th order approximate curve (B). Six-order approximation curve cookie S ₀ (A), it is a diagram illustrating a correction six linear approximation curve (B). Next six trendline cookie S _{0 '(A),} is a diagram illustrating a correction six linear approximation curve (B). Six-order approximation curve cookie S ₁ (A), it is a diagram illustrating a correction six linear approximation curve (B). Six-order approximation curve cookie S ₂ (A), it is a diagram illustrating a correction six linear approximation curve (B). Six-order approximation curve cookie S ₃ (A), it is a diagram illustrating a correction six linear approximation curve (B). It is a figure which shows the 6th approximation curve (A) of gelatin D, and a correction | amendment 6th approximation curve (B). It is a figure which shows the 6th approximation curve (A) of gelatin E, and a correction | amendment 6th approximation curve (B). It is a figure which shows the 6th approximation curve (A) of agar F, and a correction | amendment 6th approximation curve (B). It is a figure which shows the 6th approximation curve (A) of cotton tofu G, and a correction | amendment 6th approximation curve (B). It is a figure which shows the 6th approximation curve (A) of the silken tofu H, and a correction | amendment 6th approximation curve (B). It is a figure which shows the 6th approximation curve (A) of the egg tofu I, and a correction | amendment 6th approximation curve (B).

Hereinafter, the present invention will be described in more detail.
The present invention uniformly corrects a multi-order approximate curve of a strain rate-load curve of a food sample, obtains a difference between the two obtained corrected multi-order approximate curves (patterns) as an integral value, and obtains the obtained integral It is a method of judging how hard, texture, or texture the value corresponds to based on the data collected in advance, the hardness, texture, and texture of the food Is relatively evaluated between samples.
In carrying out the evaluation method of the present invention, first, the food sample (A) and the food sample (B) are pressed with a plunger, and at the same time, the load and distortion rate during pressing are continuously measured.
Here, the food sample (A) and the food sample (B) are objects for relatively evaluating differences in the hardness, texture, and texture quality of the food. It may be a sample of the same type of food. Moreover, the same sample which a certain time passed may be sufficient.

  Foods that can be evaluated in the present invention include, for example, baked confectionery such as cookies, biscuits, pies, snack confectionery, rice confectionery; gel foods such as agar, gelatin, nata de coco, tofu, konjac gel, and aloe; apple, Fruits such as pear, yellow peach, white peach, grape, blueberry, strawberry, banana, melon, watermelon, pineapple, mango, papaya; radish, turnip, carrot, pumpkin, eggplant, cherry tomato, etc. It is not limited to these foods. Of these foods, evaluation of baked confectionery such as cookies, biscuits, pies, snack confectionery, rice confectionery, and gel food such as agar, gelatin, and tofu is particularly preferable. Thus, the present invention can target a wide variety of foods.

  The size of the sample (A) and the sample (B) may be set to a size suitable for the measurement of the load and the distortion rate based on the plunger used for pressing. Moreover, the shape suitable for the measurement is not limited, and the shape of a hole, a cylinder, a rectangular parallelepiped, a cube, a sphere, and a similar shape are usually employed. For example, a cylinder having a diameter of 20 mm × a height of 2 mm to a diameter of 50 mm × a height of 8 mm; Is exemplified. In the evaluation method of the present invention, the number of samples used for evaluation is not particularly limited.

  A device that can be used to press a food sample is a device that is generally called a plunger that can perform a compression rupture test. There is no particular limitation as long as the apparatus can continuously measure the load applied during pressing (usually 0.01 to 50 mm / sec) and simultaneously the load applied during pressing and the distortion rate with respect to the load. The shape of the plunger part that presses the food sample is preferably selected in consideration of the actual chewing mode of the food to be measured. A cylindrical plunger is preferably selected. Examples of commercially available products include creep meter RE2-30005B, creep meter RE2-3305B (manufactured by Yamaden Co., Ltd., trade name), rheometer CR-500DX-S (manufactured by Rheotech Co., Ltd., trade name), and the like. However, it is not limited to these. Note that software for outputting measurement results to the outside is incorporated in these devices in advance.

For food sample (A) and food sample (B), the load and strain rate are measured continuously, and then the load that is the output data of the measuring device for each of sample (A) and sample (B). Then, based on the values of the distortion rate, the calculation is performed by the method of least squares, and a multi-order approximation curve having the distortion rate on the X axis and the load on the Y axis, for example, a distortion rate-load curve of a fifth approximation curve is created. . When the load and distortion rate, which are output data obtained by the above measurement, are approximated by the above-described multi-order equation, they can be drawn as a multi-order approximate curve that passes through substantially the center of the output data distribution. The multi-order approximate curve only needs to be a multi-order approximate curve of 5th order or higher, and usually 5th order to 6th order for convenience of calculation.
In order to create a distortion rate-load curve of this multi-order approximation curve, first, load and distortion rate data is taken into a computer as data such as spreadsheet software, for example. Here, the distortion rate (%) is a numerical value indicating how much the sample is deformed as compared with the case where no load is applied, and {(the height of the sample before applying the load−a predetermined load). The height of the sample when applied) / (the height of the sample before applying a load)} × 100 (%). For example, in actual measurement, when the height of the sample is reduced by 20%, the distortion rate is 20%. The number of times the load and strain rate are measured for one food sample varies depending on the type of food, but the total before and after the breaking point is 5 to 100 times, preferably 10 to 80 times, more preferably 10 times. -50 times can be mentioned.

  Load and distortion data captured by a computer is used to create a distortion-load curve of a fifth-order or higher order approximate curve with the X-axis distortion and the Y-axis load using the least square method. used. Specifically, an equation of a multi-order equation is obtained from load and strain rate data using the least square method, and a strain rate-load curve is created by graphing it. These can be automatically performed by using commercially available software.

  Regarding this multi-order approximation curve, for example, a fifth-order approximation distortion rate-load curve will be described in detail. Usually, there is at least one maximum value corresponding to the breaking point and subsequently at least one minimum value at which the load is most attenuated. Exists (see FIG. 1). There is usually an inflection point between each local maximum and local minimum. When the food is chewed, the tissue breaks at a certain point, that is, breaks. This break point corresponds to the maximum value that is the first peak value in the fifth-order approximation curve of the strain rate-load curve. As can be seen from FIG. 1, the relationship between the load and the strain rate when the food is chewed using the strain rate-load curve which is a fifth-order approximation curve is as the load increases from load (gf) = 0. The strain rate also increases, reaches a maximum at the breaking point, then increases until the minimum value is reached, but the load decreases, and when the minimum value is exceeded, the load increases again as the strain rate increases. Then you can explain.

  Next, the local maximum value (MaxA) and the local maximum value (MaxB) are calculated from the multi-order approximate curves created for the sample (A) and the sample (B) with the X-axis being the distortion rate and the Y-axis being the load. Ask. Many foods have one or more maximum values, but if there is only one maximum value, that value is the maximum value (MaxA), the maximum value (MaxB), and if there are multiple maximum values, The maximum value indicating the maximum load value among the maximum values is adopted as the maximum value (MaxA) and the maximum value (MaxB). As described above, the maximum value (MaxA) and the maximum value (MaxB) are considered to correspond to the breaking point.

After determining the maximum value (MaxA) and the maximum value (MaxB), the load values on the multi-order approximation curve are unified so that the load values of the maximum value (MaxA) and the maximum value (MaxB) are the same. A corrected multi-order approximate curve is generated for each of the sample (A) and the sample (B). As a result, the multi-order approximate curve can be directly compared. In order to uniformly correct the load value, for example, based on (Equation 1): (constant value) / (maximum load value having the maximum load value in a multi-order approximation curve) = correction coefficient C The correction coefficient C is calculated, and all the load values on the original multi-order approximate curve are multiplied by the correction coefficient C. The constant value may be appropriately determined in consideration of the maximum load value, and for example, 1,000 is exemplified. Then, the corrected load obtained by this means is connected for each distortion rate, and the obtained curve is used as a corrected approximate curve. In order to create the corrected approximate curve, for example, the corrected load is plotted by dividing the distortion rate by 0.01% to 2%, but the present invention is not limited to this.

After creating the corrected multi-order approximate curve, the difference of the corrected multi-order approximate curve is integrated. If the approximate shapes of the two corrected multi-order approximate curves are similar, the integral of the difference between the corrected multi-order approximate curves is small, and if the general shapes are not similar, the integral is large. That is, the degree of divergence between the two corrected multi-order approximate curves is reflected in the integration of the difference between the corrected multi-order approximate curves, and eventually represents the difference in food hardness, texture, and texture quality. For example, the integral calculation of the difference between the corrected multi-order approximate curves is performed by dividing the distortion every 0.01% to 2%, but is not limited thereto. In addition, the integration calculation is normally performed from the start point to the end point of the measurement interval, that is, throughout the measurement interval, but it is arbitrary to select the integration interval as appropriate. Here, FIG. 2 is a diagram exemplifying a difference area of the corrected multi-order approximate curve. In the example shown in FIG. 2, the integral value of the difference between the two corrected multi-order approximate curves is calculated as the area of the hatched portion between the corrected multi-order approximate curves in the integration interval of the distortion rate 0 to 85 (%). .

  EXAMPLES Hereinafter, although an Example demonstrates this invention further more concretely, this invention is not limited to the following.

Example 1
A commercially available cookie A (55 mm diameter × 8.5 mm thick cylinder) was placed on a sample stage at room temperature, and a creep meter RE2-30005B (manufactured by Yamaden Co., Ltd., trade name) was used. A load (gf) and a distortion rate (%) were measured by pressing a cylindrical plunger having a contact area of 50 mm ² from the top surface of the cookie at a speed of 1.0 mm / second. The load (gf) and strain rate (%) were measured 30 times for the same sample.
A fifth order approximate curve with the X-axis being the distortion factor and the Y-axis being the load is calculated from the measured values of the load (gf) and the distortion factor (%) obtained by the above measurement by the least square method using a computer. A strain rate-load curve was prepared. The strain rate-load curve was expressed by a quintic function formula: y = 0.0000383757x ⁵ -0.008799127x ⁴ + 0.7339137120x ³ -262.214112153x ² + 334.00712521236x + 418.9238208917. A strain rate-load curve of the fifth order approximate curve given by the above equation is shown in FIG. In the fifth-order approximation curve, the maximum value with the maximum load value was calculated, and the load value was 1788.2 gf.
Next, unified correction was performed on the load value of the fifth-order approximation curve so that the load value was 1,000 gf (in this case, the correction coefficient C described above is 0.55922). That is, the distortion coefficient-load curve of the corrected fifth-order approximate curve was created by multiplying the correction coefficient C by the load value of the entire original fifth-order approximate curve. The corrected fifth order approximate curve is shown in FIG.

Next, the same processing as the cookie A was performed on the cookie A ₂ (the same product as the cookie A, a product with a different production lot. Diameter 55 mm × thickness 8.5 mm cylinder).
A fifth order approximate curve with the X-axis being the distortion factor and the Y-axis being the load is calculated from the measured values of the load (gf) and the distortion factor (%) obtained by the above measurement by the least square method using a computer. A strain rate-load curve was prepared. The distortion-load curve was expressed by a quintic function formula: y = 0.0000414359x ⁵ −0.00936361409x ⁴ + 0.772159393387x ³ −27.4403963291x ² + 353.63803997024x + 344.9865083471. A strain rate-load curve of the fifth order approximate curve given by the above equation is shown in FIG. In the fifth-order approximate curve, when the maximum value with the maximum load value was calculated, the load value was 1820.0 gf.
Next, unified correction was performed on the load value of the fifth-order approximation curve so that the load value becomes 1,000 gf (in this case, the correction coefficient C described above is 0.54945). That is, the distortion coefficient-load curve of the corrected fifth-order approximate curve was created by multiplying the correction coefficient C by the load value of the entire original fifth-order approximate curve. The corrected fifth order approximate curve is shown in FIG.

The correction fifth order approximation curve cookies A as a standard, the integration calculation of the difference between the corrected quintic trendline Cookie A _2, ranging from the distortion rate of 0% to 70%, also, distortion of 0.1% Interval When the integrated value was calculated under the analysis conditions, 13,128 values were obtained.

(Example 2)
A commercially available biscuit B (base 60 mm × base 48 mm × thickness 9 mm cuboid) was subjected to the same treatment as in Example 1.
A fifth order approximate curve with the X-axis being the distortion factor and the Y-axis being the load is calculated from the measured values of the load (gf) and the distortion factor (%) obtained by the above measurement by the least square method using a computer. A strain rate-load curve was prepared. The distortion-load curve was expressed by a quintic function formula: y = 0.0000478682x ⁵ -0.0108325413x ⁴ + 0.8968807150x ³ -32.0799328044 x ² + 4211.74231016088 + 35.1647268115. FIG. 5A shows a distortion rate-load curve of the fifth order approximate curve given by the above equation. In the fifth-order approximation curve, the maximum value with the maximum load value was calculated, and the load value was 1839.4 gf.
Next, unified correction was performed on the load value of the fifth-order approximation curve so that the load value becomes 1,000 gf (in this case, the correction coefficient C described above is 0.54365). That is, the distortion coefficient-load curve of the corrected fifth-order approximate curve was created by multiplying the correction coefficient C by the load value of the entire original fifth-order approximate curve. The corrected fifth order approximate curve is shown in FIG.

The integral calculation of the difference from the corrected fifth-order approximate curve of cookie B, using the corrected fifth-order approximate curve of cookie A shown in Example 1 as a standard, is performed in the range of the distortion rate from 0% to 70%. When integrated values were calculated under analysis conditions of 0.1% intervals, values of 21,542 were obtained.

(Example 3)
A commercially available biscuit C (diameter 60 mm × thickness 6 mm cylinder) was treated in the same manner as in Example 1.
A fifth order approximate curve with the X-axis being the distortion factor and the Y-axis being the load is calculated from the measured values of the load (gf) and the distortion factor (%) obtained by the above measurement by the least square method using a computer. A strain rate-load curve was prepared. The strain rate-load curve was expressed by a quintic function formula: y = 0.0003000152x ⁵ -0.0067157687x ⁴ + 0.57488110521x ³ -23.4186267445x ² + 412.41187858289x-549.4355481123. FIG. 6A shows a distortion rate-load curve of the fifth order approximate curve given by the above equation. In the fifth-order approximate curve, when the maximum value with the maximum load value was calculated, the load value was 2001.8 gf.
Next, unified correction was performed on the load value of the fifth-order approximation curve so that the load value becomes 1,000 gf (in this case, the correction coefficient C described above is 0.49955). That is, the distortion coefficient-load curve of the corrected fifth-order approximate curve was created by multiplying the correction coefficient C by the load value of the entire original fifth-order approximate curve. The corrected fifth order approximate curve is shown in FIG.

The integral calculation of the difference from the corrected fifth-order approximate curve of cookie C, using the corrected fifth-order approximate curve of cookie A shown in Example 1 as a standard, is performed in the range from 0% to 70% of the distortion rate. When the integration value was calculated under the analysis conditions of 0.1% intervals, 255 and 680 values were obtained.

(Evaluation of Examples 1 to 3)
In sensory evaluation, Cookie A and Cookie A ₂ have the same texture, biscuits B have a texture similar to Cookie A, and biscuits C have a texture that is completely different from Cookie A. Has been.
On the other hand, the integral value of the difference obtained in Example 1 regarding Cookie A and Cookie A ₂ was 13,128, and the value was the smallest among Examples 1 to 3. The integrated value of the difference obtained in Example 2 for Cookie A and Biscuit B was 21,542, which was small. The integrated value of the difference obtained in Example 3 for Cookie A and Biscuit C was 255,680, which was large.
As can be seen from this result, the integrated values obtained in Examples 1 to 3 accurately reflected the actual sensory evaluation results.

Example 4
The fracture curve approximation of the measurement result of cookie A shown in Example 1 was changed from the fifth-order equation approximation to the sixth-order equation approximation and analyzed. The distortion-load curve was represented by a sixth-order function formula: y = −0.0000000584790x ⁶ + 0.000196064433x ⁵ −0.024898944982x ⁴ + 1.5044695292018x ³ −43.4480801118127x ² + 488.1947093497773x + 96.927740182060 A distortion rate-load curve of the sixth-order approximation curve given by the above equation is shown in FIG. In the sixth-order approximation curve, when the maximum value with the maximum load value was calculated, the load value was 1913.7 gf.
Next, unified correction was performed on the load value of the sixth approximation curve so that the load value becomes 1,000 gf (in this case, the correction coefficient C is 0.52254). That is, the distortion factor-load curve of the corrected sixth-order approximate curve was created by multiplying the correction coefficient C by the load value of the entire original sixth-order approximate curve. The corrected sixth order approximate curve is shown in FIG.

Subsequently, the approximation of the break curve for the measurement result of cookie A ₂ shown in Example 1 was changed from the quintic equation approximation to the sixth equation approximation and analyzed. The strain rate-load curve was represented by a sixth-order function formula: y = −0.00000000260260x ⁶ + 0.00011162367x ⁵ −0.0165529353287x ⁴ +1.1152797575858 x ³ −35.128535536380x ² + 422.2257769022195x + 201.634992668681 A distortion rate-load curve of the sixth-order approximation curve given by the above equation is shown in FIG. In the sixth-order approximation curve, the maximum value with the maximum load value was calculated, and the load value was 1873.2 gf.
Next, unified correction was performed on the load value of the sixth approximation curve so that the load value becomes 1,000 gf (in this case, the correction coefficient C is 0.53385). That is, the distortion factor-load curve of the corrected sixth-order approximate curve was created by multiplying the correction coefficient C by the load value of the entire original sixth-order approximate curve. The corrected sixth order approximate curve is shown in FIG.

A correction next six trendline Cookie A as a standard, the integration calculation of the difference between the corrected six-order approximation curve cookies A _2, ranging from the distortion rate of 0% to 70%, also, distortion of 0.1% Interval When the integrated value was calculated under the analysis conditions, a value of 22,000 was obtained.

(Example 5)
The fracture curve approximation for the measurement result of biscuit B shown in Example 2 was changed from the quintic equation approximation to the sixth equation approximation and analyzed. The distortion-load curve was expressed by a sixth-order function formula: y = −0.0000000269589x ⁶ + 0.000120635174x ⁵ −0.01826815597x ⁴ + 1.2530538850107x ³ −40.067130250505x ² + 492.9707077931471x-11.057555575364 FIG. 9A shows a distortion rate-load curve of a sixth-order approximation curve given by the above equation. In the sixth-order approximation curve, when the maximum value with the maximum load value was calculated, the load value was 1892.3 gf.
Next, unified correction was performed on the load value of the sixth approximation curve so that the load value becomes 1,000 gf (in this case, the correction coefficient C described above is 0.52844). That is, the distortion factor-load curve of the corrected sixth-order approximate curve was created by multiplying the correction coefficient C by the load value of the entire original sixth-order approximate curve. The corrected sixth order approximate curve is shown in FIG.

The integral calculation of the difference from the corrected sixth-order approximate curve of cookie A shown in Example 4 and the correction sixth-order approximate curve of biscuit B is performed in the range from 0% to 70%, and the distortion rate. When integrated values were calculated under analysis conditions at 0.1% intervals, values of 60,197 were obtained.

(Example 6)
The fracture curve approximation for the measurement result of the biscuit C shown in Example 3 was changed from the fifth-order equation approximation to the sixth-order equation approximation and analyzed. The distortion-load curve was represented by a sixth-order function formula: y = 0.00000476258x ⁶ -0.0000976633788x ⁵ + 0.006422710726x ⁴ -0.0543882492102x ³ -9.3334177475618x ² + 287.202060697600x-292.1157341497011. FIG. 10A shows a distortion rate-load curve of the sixth-order approximation curve given by the above formula. In the sixth-order approximate curve, when the maximum value with the maximum load value was calculated, the load value was 2042.2 gf.
Next, unified correction was performed on the load value of the sixth approximation curve so that the load value was 1,000 gf (in this case, the correction coefficient C described above is 0.48967). That is, the distortion factor-load curve of the corrected sixth-order approximate curve was created by multiplying the correction coefficient C by the load value of the entire original sixth-order approximate curve. The corrected sixth order approximate curve is shown in FIG.

The integral calculation of the difference from the corrected sixth-order approximation curve of cookie A shown in Example 4 and the correction sixth-order approximation curve of biscuit C is performed in the range from 0% to 70% distortion. When integrated values were calculated under analysis conditions of 0.1% intervals, values of 269,068 were obtained.

(Evaluation of Examples 4 to 6)
The integrated value of the difference obtained in Example 4 with respect to Cookie A and Cookie A ₂ was 22,000, and the value among Examples 4 to 6 was the smallest. The integrated value of the difference obtained in Example 5 for Cookie A and Biscuit B was 60,197, which was small. The integrated value of the difference obtained in Example 6 for Cookie A and Biscuit C was 269,068, which was large.
As can be seen from this result, the integrated values obtained in Examples 4 to 6 accurately reflected the actual sensory evaluation results.
As described above, it is possible to evaluate the quality of the texture with the integral value of the difference in the sixth-order equation approximation as well as the fifth-order equation approximation.

(Example 7)
For the cookie A shown in Example 1 immediately after opening (hereinafter referred to as “cookie S ₀ ”), the same processing as in Example 1 was performed, and the distortion-load curve for the measurement result was approximated to a sixth-order equation. And analyzed. The distortion-load curve was represented by a sixth-order function formula: y = −0.0000000584790x ⁶ + 0.000196064433x ⁵ −0.024898944982x ⁴ + 1.5044695292018x ³ −43.4480801118127x ² + 488.1947093497773x + 96.927740182060 FIG. 11A shows a distortion rate-load curve of a sixth-order approximation curve given by the above equation. In the sixth-order approximation curve, when the maximum value with the maximum load value was calculated, the load value was 1913.7 gf.
Next, unified correction was performed on the load value of the sixth-order approximation curve so that the load value was 1,000 gf (in this case, the correction coefficient C described above is 0.52254). That is, the distortion factor-load curve of the corrected sixth-order approximate curve was created by multiplying the correction coefficient C by the load value of the entire original sixth-order approximate curve. The corrected sixth order approximate curve is shown in FIG.

For the cookie A shown in Example 1 immediately after opening and the same product lot different product (hereinafter referred to as “cookie S ₀ ′”), the same processing as in Example 1 was performed, and the distortion rate-load for the measurement result The curve was approximated and analyzed by a sixth-order equation. The strain rate-load curve was represented by a sixth-order function formula: y = −0.00000000260260x ⁶ + 0.00011162367x ⁵ −0.0165529353287x ⁴ +1.1152797575858 x ³ −35.128535536380x ² + 422.2257769022195x + 201.634992668681 FIG. 12A shows a distortion rate-load curve of the sixth-order approximation curve given by the above formula. In the sixth-order approximation curve, the maximum value with the maximum load value was calculated, and the load value was 1873.2 gf.
Next, unified correction was performed on the load value of the sixth approximation curve so that the load value was 1,000 gf (in this case, the correction coefficient C described above is 0.53385). That is, the distortion factor-load curve of the corrected sixth-order approximate curve was created by multiplying the correction coefficient C by the load value of the entire original sixth-order approximate curve. The corrected sixth order approximate curve is shown in FIG.

A correction next six trendline cookie S ₀ as a standard, the integration calculation of the difference between the corrected six-order approximation curve cookie S _{0 ',} ranging from the distortion rate of 0% to 70%, also, distortion 0.1 When the integrated value was calculated under the analysis conditions of% intervals, a value of 22,000 was obtained.

(Example 8)
About cookies S, temperature 20 ℃, was subjected to a moisture abuse one day at a humidity of 60% of the constant temperature and humidity chamber, the cookies S after one day opening a cookie S _1, the approximate six equations breaking curve for the measurement result And analyzed. The distortion-load curve was expressed by a sixth-order function formula: y = −0.00000003953313x ⁶ + 0.0001546665504x ⁵ −0.021838118211x ⁴ + 1.428950883531x ³ −43.9999744926266x ² + 526.8890794571256x−86.780080367811 FIG. 13A shows a distortion rate-load curve of a sixth-order approximation curve given by the above equation. In the sixth-order approximation curve, when the maximum value with the maximum load value was calculated, the load value was 2008. 1gf.
Next, unified correction was performed on the load value of the sixth approximation curve so that the load value was 1,000 gf (in this case, the correction coefficient C described above is 0.494797). That is, the distortion factor-load curve of the corrected sixth-order approximate curve was created by multiplying the correction coefficient C by the load value of the entire original sixth-order approximate curve. The corrected sixth-order approximate curve is shown in FIG.

Using the corrected sixth-order approximate curve of cookie S ₀ shown in Example 7 as a standard, the integral calculation of the difference from the corrected sixth-order approximate curve of cookie S ₁ is performed in the range from 0% to 70% distortion, When integrated values were calculated under analysis conditions with a distortion rate of 0.1% intervals, 37,303 values were obtained.

Example 9
Cookies S, temperature 20 ℃, was subjected to a moisture abuse for 2 days at a humidity of 60% of the constant temperature and humidity chamber, the cookies S 2 days after opening a cookie S _2, and approximate six equations breaking curve for the measurement result And analyzed. The distortion-load curve was represented by a sixth-order function formula: y = −0.00000000161969x ⁶ + 0.000103176285x ⁵ −0.011818466472x ⁴ + 1.3766031926282x ³ −48.017703747610x ² + 671.5316554559721x−590.9200640115888 FIG. 14A shows a distortion rate-load curve of a sixth-order approximation curve given by the above formula. In the sixth-order approximate curve, when the maximum value with the maximum load value was calculated, the load value was 2578.8 gf.
Next, unified correction was performed on the load value of the sixth approximation curve so that the load value becomes 1,000 gf (in this case, the correction coefficient C described above is 0.38777). That is, the distortion factor-load curve of the corrected sixth-order approximate curve was created by multiplying the correction coefficient C by the load value of the entire original sixth-order approximate curve. The corrected sixth-order approximate curve is shown in FIG.

A correction next six trendline cookie S ₀ shown in Example 7 as the standard, the integration calculation of the difference between the corrected six-order approximation curve cookies S _2, in the range of from strain of 0% to 70%, and, When integrated values were calculated under analysis conditions with a distortion rate of 0.1% interval, values of 133 and 528 were obtained.

(Example 10)
Cookies S, temperature 20 ℃, was subjected to a 3-day moisture abuse at a humidity of 60% of the constant temperature and humidity chamber, the cookies S of three days after opening and cookies S _3, and approximate six equations breaking curve for the measurement result And analyzed. The distortion-load curve was expressed by a sixth-order function formula: y = 0.0000000760962x ⁶ -0.0001762227065x ⁵ + 0.014608809896x ⁴ -0.467809894535x ³ + 0.5480078750302x ² + 227.7745507761695x-330.827401001006. FIG. 15A shows a distortion-load curve of a sixth-order approximation curve given by the above formula. In the sixth-order approximation curve, when the maximum value with the maximum load value was calculated, the load value was 2573.3 gf.
Next, unified correction was performed on the load value of the sixth approximation curve so that the load value becomes 1,000 gf (in this case, the correction coefficient C described above is 0.38861). That is, the distortion factor-load curve of the corrected sixth-order approximate curve was created by multiplying the correction coefficient C by the load value of the entire original sixth-order approximate curve. The corrected sixth order approximate curve is shown in FIG.

And a correction next six trendline cookie S ₀ shown in Example 7 as the standard, the integration calculation of the difference between the corrected six-order approximation curve cookies S _3, in a range from the strain of 0% to 70%, and, When integrated values were calculated under analysis conditions with a distortion rate of 0.1% interval, values of 388 and 330 were obtained.

(Evaluation of Examples 7 to 10)
In Examples 7 to 10, assuming the application of the present invention to food quality control, the method of the present invention was used to track changes in texture with respect to moisture of cookies as a model.
According to the sensory evaluation, cookies S ₀ and the product is a standard, and is a cookie S _{0 'is} production lot differences article, at all has a texture homogeneous, slightly in comparison to the cookies S ₁ Cookies S ₀ to have texture varying degrees feel moisture, comparison of the comparison cookies S ₂ cookies S _0, there is texture changes in the extent to which crispness is lost clearly, the cookie S ₃ cookies S ₀ It has been evaluated that the crispness has been completely lost and has changed to a heavy texture.
On the other hand, the integrated value of the difference obtained in Example 7 regarding cookie S ₀ and cookie S ₀ ′ was 22,000, and was the smallest among Examples 7-11. The integrated value of the difference obtained in Example 8 for cookie S ₀ and cookie S ₁ was 37,303, which was small on the first day of opening. The integrated value of the difference obtained in Example 9 with respect to cookie S ₀ and cookie S ₂ was 133,528, and an increase in value was observed on the second day of opening. The integrated value of the difference obtained in Example 10 for cookie S ₀ and cookie S ₃ was 388,330, and the value was very large on the third day of opening.
As can be seen from this result, the integrated values obtained in Examples 7 to 10 accurately reflected the actual sensory evaluation results.
Thus, it was found that the present invention can be applied to baked confectionery for the purpose of evaluating the change in texture due to moisture.

(Example 11)
A commercially available gelatin D (bottom 10 mm × bottom 10 mm × thickness 10 mm cube) was placed on a sample stage with a product temperature of 4
The columnar plunger with a contact area of 50 mm ² was placed at a speed of 1.0 mm / sec from the top surface of the gelatin using a creep meter RE2-30005B (trade name, manufactured by Yamaden Co., Ltd.). The load (gf) and the strain rate (%) were measured by pressing with. The load (gf) and strain rate (%) were measured 30 times for the same sample.
A sixth-order approximate curve with the X-axis being the distortion factor and the Y-axis being the load is calculated from the measured values of the load (gf) and the distortion factor (%) obtained by the above measurement by the least square method using a computer. A strain rate-load curve was prepared. The distortion-load curve was expressed by a sixth-order function formula: y = 0.000000016752x ⁶ -0.0000029111588x ⁵ + 0.000122187228x ⁴ + 0.0018996019425x ³ -0.1443437709116x ² + 2.0283737387700x-2.5777150862664 FIG. 16A shows a distortion rate-load curve of a sixth-order approximation curve given by the above equation. In the sixth-order approximate curve, when the maximum value with the maximum load value was calculated, the load value was 110.1 gf.
Next, unified correction was performed on the load value of the sixth-order approximate curve so that the load value was 1,000 gf (in this case, the correction coefficient C is 9.08554). That is, the distortion factor-load curve of the corrected sixth-order approximate curve was created by multiplying the correction coefficient C by the load value of the entire original sixth-order approximate curve. The corrected sixth order approximate curve is shown in FIG.

Commercially available gelatin E (bottom 10 mm × bottom 10 mm × thickness 10 mm cube) was processed in the same manner as gelatin D above.
A sixth-order approximate curve with the X-axis being the distortion factor and the Y-axis being the load is calculated from the measured values of the load (gf) and the distortion factor (%) obtained by the above measurement by the least square method using a computer. A strain rate-load curve was prepared. Its distortion - load curve, six order function ^formula expressed by ^{y = 0.000000053977x 6 -0.000013689109x 5 + 0.001284817030x} 4 -0.055535504777x 3 + 1.142779167087x 2 -8.946977781481x + 17.650545045704. FIG. 17A shows a distortion rate-load curve of the sixth-order approximation curve given by the above equation. In the sixth-order approximation curve, when the maximum value with the maximum load value was calculated, the load value was 136.3 gf.
Next, unified correction was performed on the load value of the sixth approximation curve so that the load value becomes 1,000 gf (in this case, the correction coefficient C is 7.33786). That is, the distortion factor-load curve of the corrected sixth-order approximate curve was created by multiplying the correction coefficient C by the load value of the entire original sixth-order approximate curve. The corrected sixth order approximate curve is shown in FIG.

The integral calculation of the difference from the corrected sixth-order approximate curve of gelatin E with the corrected sixth-order approximate curve of gelatin D as a standard is performed in the range from 0% to 70% distortion, and at intervals of 0.1% distortion. When the integration value was calculated under the analysis conditions, 67,858 values were obtained.

(Example 12)
A commercially available agar F (bottom 10 mm × bottom 10 mm × thickness 10 mm cube) was treated in the same manner as in Example 11.
A sixth-order approximate curve with the X-axis being the distortion factor and the Y-axis being the load is calculated from the measured values of the load (gf) and the distortion factor (%) obtained by the above measurement by the least square method using a computer. A strain rate-load curve was prepared. The strain rate-load curve was expressed by a sixth-order function formula: y = 0.000000068333x ⁶ -0.0000210354464x ⁵ + 0.002476495285x ⁴ -0.137178869942x ³ + 3.4595967737710x ² -28.190001217736x + 54.0413708813822. FIG. 18A shows a distortion-load curve of a sixth-order approximation curve given by the above formula. In the sixth-order approximation curve, when the maximum value with the maximum load value was calculated, the load value was 160.3 gf.
Next, unified correction was performed on the load value of the sixth approximation curve so that the load value becomes 1,000 gf (in this case, the correction coefficient C is 6.24017). That is, the distortion factor-load curve of the corrected sixth-order approximate curve was created by multiplying the correction coefficient C by the load value of the entire original sixth-order approximate curve. The corrected sixth order approximate curve is shown in FIG.

Using the corrected sixth-order approximate curve of gelatin D shown in Example 11 as a standard, the integral calculation of the difference from the corrected sixth-order approximate curve of agar F is performed in the range of the distortion rate from 0% to 70%. When integrated values were calculated under analysis conditions of 0.1% intervals, 260 and 391 values were obtained.

(Example 13)
A commercially available cotton tofu G (bottom 10 mm × bottom 10 mm × thickness 10 mm cube) was subjected to the same treatment as in Example 11.
A sixth-order approximate curve with the X-axis being the distortion factor and the Y-axis being the load is calculated from the measured values of the load (gf) and the distortion factor (%) obtained by the above measurement by the least square method using a computer. A strain rate-load curve was prepared. The distortion-load curve was expressed by a sixth-order function formula: y = −0.000000000002563x ⁶ + 0.000001170355x ⁵ -0.0001525598958x ⁴ + 0.0074978843979x ³ -0.130113219928x ² + 1.65119621156x + 0.2798151108574. FIG. 19A shows a distortion rate-load curve of a sixth-order approximation curve given by the above formula. In the sixth-order approximation curve, the maximum value with the maximum load value was calculated, and the load value was 67.1 gf.
Next, unified correction was performed on the load value of the sixth approximation curve so that the load value becomes 1,000 gf (in this case, the correction coefficient C is 14.8943). That is, the distortion factor-load curve of the corrected sixth-order approximate curve was created by multiplying the correction coefficient C by the load value of the entire original sixth-order approximate curve. The corrected sixth-order approximate curve is shown in FIG.

Using the corrected sixth-order approximate curve of gelatin D shown in Example 11 as a standard, the integral calculation of the difference from the corrected sixth-order approximate curve of cotton tofu G is performed in the range from 0% to 70% distortion, When integrated values were calculated under analysis conditions with a rate of 0.1% intervals, values of 137 and 858 were obtained.

(Example 14)
A commercially available silken tofu H (bottom 10 mm × bottom 10 mm × thickness 10 mm cube) was subjected to the same treatment as in Example 11.
A sixth-order approximate curve with the X-axis being the distortion factor and the Y-axis being the load is calculated from the measured values of the load (gf) and the distortion factor (%) obtained by the above measurement by the least square method using a computer. A strain rate-load curve was prepared. The distortion-load curve was expressed by a sixth-order function formula: y = 0.000000017437x ⁶ -0.000004846631x ⁵ + 0.000514129432x ⁴ -0.025520889205x ³ + 0.5675339614379x ² -3.221335041217x + 7.22533306625033 A distortion rate-load curve of a sixth-order approximation curve given by the above equation is shown in FIG. In the sixth-order approximation curve, when the maximum value with the maximum load value was calculated, the load value was 44.9 gf.
Next, unified correction was performed on the load value of the sixth approximation curve so that the load value becomes 1,000 gf (in this case, the correction coefficient C is 22.2832). That is, the distortion factor-load curve of the corrected sixth-order approximate curve was created by multiplying the correction coefficient C by the load value of the entire original sixth-order approximate curve. The corrected sixth order approximate curve is shown in FIG.

Using the corrected sixth-order approximate curve of gelatin D shown in Example 11 as a standard, the integral calculation of the difference from the corrected sixth-order approximate curve of silken tofu H was performed in the range from 0% to 70% distortion, When integrated values were calculated under analysis conditions with a rate of 0.1% intervals, values of 283 and 227 were obtained.

(Example 15)
A commercially available egg tofu I (bottom 10 mm × bottom 10 mm × thickness 10 mm cube) was processed in the same manner as in Example 11.
A sixth-order approximate curve with the X-axis being the distortion factor and the Y-axis being the load is calculated from the measured values of the load (gf) and the distortion factor (%) obtained by the above measurement by the least square method using a computer. A strain rate-load curve was prepared. The distortion-load curve was represented by a sixth-order function formula: y = 0.0000000002270x ⁶ -0.0000000639133x ⁵ + 0.0000719644588x ⁴ -0.0040214889511x ³ + 0.100948442288x ² -0.1640407488849x + 1.571008583218 A distortion rate-load curve of the sixth-order approximation curve given by the above equation is shown in FIG. In the sixth-order approximate curve, when the maximum value with the maximum load value was calculated, the load value was 28.6 gf.
Next, unified correction was performed on the load value of the sixth-order approximation curve so that the load value was 1,000 gf (in this case, the correction coefficient C is 34.9578). That is, the distortion factor-load curve of the corrected sixth-order approximate curve was created by multiplying the correction coefficient C by the load value of the entire original sixth-order approximate curve. The corrected sixth-order approximate curve is shown in FIG.

Using the corrected sixth-order approximation curve of gelatin D shown in Example 11 as a standard, the integral calculation of the difference from the corrected sixth-order approximation curve of egg tofu I is performed in the range of distortion rate from 0% to 70%, When integrated values were calculated under analysis conditions with a rate of 0.1% intervals, 201 and 231 values were obtained.

(Evaluation of Examples 11-15)
In Examples 11-15, the method of this invention was provided with respect to the gel-like foodstuff.
According to sensory evaluation, the standard gelatin D and gelatin E are different products, but the texture quality is the same texture, which is the closest of the samples prepared here. Silken tofu G and egg tofu I have a texture different from that of standard gelatin D, and cotton tofu G is evaluated to have a slightly similar texture to gelatin D. .
On the other hand, the integrated value of the difference obtained in Example 11 with respect to gelatin D and gelatin E was 67,858, which was the smallest among Examples 11-15. The integrated values of the differences obtained in Examples 12, 14, and 15 for gelatin D, agar F, silken tofu G, and egg tofu I were 260, 391, 283, 227, 201, and 231 respectively, and the values were slightly large. . The integrated value of the difference obtained in Example 13 for cotton tofu G was 137,858, which was a medium value.
As can be seen from the results, the integrated values obtained in Examples 11 to 15 accurately reflected the actual sensory evaluation results.
Thus, the quality of the texture could be evaluated by applying the present invention not only to baked confectionery but also to gel food.

Claims (2)

A food sample (A) is pressed with a plunger, and simultaneously the load and strain rate during pressing are measured continuously. Based on the values of the load and strain rate, calculation is performed by the method of least squares, and X Creating a strain-one-load curve of a fifth-order or higher order approximate curve with the axis as the strain rate and the Y-axis as the load, and calculating the maximum value (MaxA) in the multi-order approximate curve,
On the other hand, for the food sample (B), in the same manner as described above, a strain rate-one load curve of a homogeneous approximation curve is created, and the maximum value (MaxB) in the approximation curve is calculated and obtained.
Next, a corrected multi-order approximate curve obtained by uniformly correcting the load values on the multi-order approximate curve so that the load values of the maximum value (Max A) and the maximum value (Max B) are the same value is obtained as the multi-order approximate curve. Create for each of the fitted curves,
Next, the difference between the created two corrected multi-order approximate curves is integrated, and using the obtained integrated value, the difference in food texture between the food sample (A) and the food sample (B) is evaluated. Relative evaluation method by pattern recognition of food texture. The relative evaluation method according to claim 1, wherein the multi-order approximate curve is a fifth-order approximate curve or a sixth-order approximate curve.

Images