Meat identification method, method for constructing database for meat identification, and meat identification system

Abstract

To provide a meat discriminating method capable of finely discriminating non-meat parts such as skin, bone, fat and hair and specifying the positions of these parts.SOLUTION: The method includes a step of irradiating meat with light while changing a wavelength of the light, a step of acquiring reflectance data of each of determination parts of the meat from a reflectance for each wavelength, a step of selecting at least one wavelength at which a difference in reflectance between the determination parts is larger than those of the other wavelengths from the reflectance data and the wavelengths, a step of setting a determination region from a distribution of the reflectances of the determination parts at the selected wavelength, and a step of storing the determination region for each of the determination parts in a storage unit. The step of performing meat discrimination includes a step of irradiating the entire meat with light of at least one wavelength, and a step of obtaining a reflectance from reflected light and specifying a position of the reflected light belonging to which discrimination region.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present invention relates to a meat discrimination method capable of discriminating between the lean meat, skin, fat, hair, bones, etc. in the meaty parts of meat and other non-meaty parts, a method for building a database for meat discrimination, and a meat discrimination system using this database. [Background technology]

[0002] For example, yakitori is a dish in which chicken meat cut into specified sizes is skewered and grilled. When processing chicken for yakitori, it is necessary to consider the balance of fat and lean meat, as well as the balance of skin and fat, before cutting it. Chicken skin has fat and thick hair attached to it. In the processing process, the parts to be distinguished—skin, fat, and hair—are first identified and the hair is removed. After the excess fat is trimmed off, the chicken skin and lean meat are cut into specified sizes. When processing chicken for yakitori, workers visually check the chicken skin for excess fat and hair and cut them off.

[0003] However, such manual work is inefficient, with large differences in work efficiency depending on the level of skill, the burden on workers is heavy because they have to work in the same position for long periods of time, and a high level of care is required to prevent the contamination and proliferation of germs by manual labor. For these reasons, it is desirable to automate some or all of this work as much as possible.

[0004] Hyperspectral imaging (HSI), which is used to detect foreign objects in food, is known as a method for distinguishing between meaty and non-meaty parts (including lean meat, skin, fat, bone, and hair) (see, for example, Patent Document 1). HSI can obtain spectral information for each wavelength from ultraviolet to visible light and infrared, and has the advantage of being able to nondestructively distinguish between materials that are difficult to distinguish with the naked eye or with an RGB camera, such as chicken skin. However, HSI has the disadvantage that it cannot measure three-dimensional objects such as chicken.

[0005] Patent Document 2 discloses a method for identifying bones in meat, which involves irradiating a cut surface of meat with visible light and receiving the reflected light to obtain a first image, irradiating the cut surface with infrared light having a wavelength of 1100 nm to 1700 nm and receiving the reflected light to obtain a second image, and calculating the difference in brightness between the first and second images to extract only the bone area. The discrimination method described in this document can discriminate bone regions exposed on the cut surface of meat with high accuracy, but has the problem that it is difficult to discriminate bone regions in meat with a three-dimensional shape, and is not suitable for fine discrimination of lean meat, skin, bones, fat, hair, etc. in meat. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Japanese Patent Application Laid-Open No. 2024-75104 [Patent Document 2] Japanese Patent Application Laid-Open No. 2018-66649 DISCLOSURE OF THE INVENTION [Problem to be solved by the invention]

[0007] The present invention aims to solve the problems of the conventional techniques described above, and to provide a meat discrimination method, a method for building a database for meat discrimination, and a meat discrimination system that can distinguish between discriminable parts such as lean meat, skin, bones, fat, and hair in the meaty and non-meaty parts of meat in detail, can be applied to meat with a three-dimensional shape, and can also identify the positions of discriminable parts (including parts with a certain extent) in the meaty and non-meaty parts. [Means for solving the problem]

[0008] In order to solve the above problem, the invention of claim 1 is a meat discrimination method for discriminating discriminable parts in meat parts and / or non-meaty parts in a non-contact manner by irradiating light onto meat, and the method comprises a preparation step of collecting data necessary for meat discrimination, and a discrimination step of discriminating discriminable parts in meat parts and / or non-meaty parts of the meat, the preparation step includes a step of preparing a light source that can irradiate the meat with light at different wavelengths and a light receiving means that receives reflected light from the meat, a step of changing the wavelength of the irradiated light for each type of meat part and each type of non-meaty part, and irradiating at least a part of the meat with the light, a step of obtaining reflectance data for each discriminable part in the meat parts and / or non-meaty parts of the meat from the reflectance of the meat parts and / or the reflectance of the non-meaty parts for each wavelength obtained by the light irradiation, and a step of calculating the reflectance of each discriminable part from the reflectance data and the wavelength. the step of selecting at least one wavelength at which the difference between the wavelengths is larger than the others, the step of setting a discrimination area from the distribution of reflectance for each of the discriminated parts at the selected wavelengths in order to discriminate each of the discriminated parts, and the step of storing the discrimination area for each of the discriminated parts in memory means, wherein the discrimination step of performing meat discrimination comprises the steps of preparing a light source that can irradiate the meat with light of at least one wavelength set in the preparation step and light receiving means that receives reflected light from the meat, irradiating the entire meat with light of at least one of the wavelengths, determining the reflectance from the reflected light of the irradiated light, and comparing it with the discrimination area read out from the memory means to determine which of the discrimination areas of the discriminated part the reflectance belongs to, and the step of identifying the position of the reflected light that belongs to any of the discrimination areas and displaying the distribution of each of the discriminated parts of the meat.

[0009] As described in claim 2, the discrimination area may be set from reflectance data for a plurality of different wavelengths, or as described in claim 3, the meat may be measured in three dimensions, and the discriminated portions of the meaty portion and the non-meaty portion may be discriminated from the results of this three-dimensional measurement and the reflectance. The light of the selected wavelength may be irradiated and scanned over the entire meat, and the position of the scanned area may be identified from the reflection angle of the reflected light, thereby simultaneously performing the three-dimensional measurement and distinguishing between the meaty and non-meaty areas based on the reflectance.

[0010] As described in claim 5, the thickness of the non-meaty portion may be obtained from the results of the three-dimensional measurement, and when the thickness of the non-meaty portion obtained and the type of the non-meaty portion exceed a predetermined threshold, it may be determined that different types of non-meaty portions are layered. As described in claim 6, the discrimination method of the present invention can be suitably used for chicken meat, and can discriminate fat, skin, or hair attached to the chicken meat. In this case, as described in claim 7, the discrimination region may be set based on the reflectance at wavelengths of 1200±50 nm and 1550±50 nm.

[0011] The method for constructing a database for meat discrimination of the present invention, as recited in claim 8, comprises the steps of preparing a light source that can irradiate the meat with light at different wavelengths and light-receiving means that receives light reflected from the meat, and irradiating at least a part of the meat with light of different wavelengths for each type of meat part and each type of non-meaty part, obtaining reflectance data for each of the distinguished parts of the meat from the reflectances of the meat parts and the non-meaty parts for each wavelength obtained by the light irradiation, selecting at least one wavelength at which the difference in reflectance of each of the distinguished parts is larger than the others, from the reflectance data of the meat part or the non-meaty part and the wavelength, setting a discrimination region for each of the distinguished parts from the distribution of reflectance for each of the distinguished parts at the selected wavelength, and storing the discrimination region for each of the distinguished parts in storage means. As described in claim 9, the discrimination region may be set based on reflectance data for a plurality of different wavelengths.

[0012] As described in claim 10, the method may further include the steps of: storing the thickness of the non-meaty part as thickness data from the results obtained by three-dimensional measurement of the meat; and determining and storing a threshold value of thickness to be used for determining, from the thickness data and the type of the non-meaty part, that a non-discriminatable part is layered on top of a discriminated part of the non-meaty part. The meat discrimination system of the present invention, as set forth in claim 11, is a meat discrimination system that distinguishes between meaty parts and non-meaty parts in a non-contact manner by irradiating meat with light, and comprises a light source that irradiates light of one or more wavelengths predetermined according to the type of meat, irradiation means that irradiates at least a part of the meat with the light, light-receiving means that receives the reflected light of the light, reflectance measuring means that determines the reflectance from the reflected light received by the light-receiving means, and calculation means that is connected to the database by a communication line, reads out a discrimination area for each of the distinguished parts according to the type of meat, determines which of the distinguished parts the measured reflectance of light belongs to as the discrimination area, and identifies where in the meat the reflectance is located from the results of three-dimensional measurement of the meat. BEST MODE FOR CARRYING OUT THE INVENTION

[0013] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the drawings. Figure 1 is a photograph showing the skin (chicken skin), fat, and hair, which are the non-meaty parts of chicken meat to be distinguished in this embodiment, using chicken as the type of meat; Figure 2 is a schematic diagram of a data collection device that collects measurement results for each type of meat as data, creates data for meat discrimination, and stores it in a database; Figure 3 is a flow chart explaining the data collection procedure; Figure 4 is a graph for each sample showing the relationship between wavelength and reflectance, obtained by calculating the reflectance measurement results; and Figure 5 is a diagram showing the distribution of reflectance for each sample at two wavelengths selected based on the results of the calculation in Figure 4. As shown in FIG. 1, the parts to be identified in the chicken meat in this embodiment are the skin, fat, and hair on the surface of the skin.

[0014] [Data collection device] As shown in Figure 2, the data collection device 1 for building a database for meat discrimination has as its main component a data processing device 10, and has a light source 11 that generates light of a preset wavelength, an irradiation unit 12 that irradiates the meat with light of the specified wavelength generated by this light source 11, a photodetector 13 that receives reflected light from the irradiated area, a reflectance measuring device 14 that measures the reflectance, which is the ratio of the reflected light to the irradiated light, from the reflected light received by the photodetector 13, and a database DB for storing data processed by the data processing device 10. The data processing device 10 includes a processor such as a CPU or MPU, memory, storage, and a communication interface, and can be a commercially available general-purpose personal computer. The data processing device 10, the laser light source 11, the reflectance measuring device 14, and the database DB are connected to each other via a bus so that they can communicate with each other.

[0015] The light source 11 may be any light source capable of irradiating the target portion of meat with light of a predetermined wavelength and distinguishing the target portion in the non-meaty portion based on the difference in reflectance, and may be capable of tunable wavelength within a certain range. The photodetector 13 may be any light source capable of receiving reflected light whose reflectance can be measured by the reflectance measuring device 14. However, when light with wavelengths ranging from visible light to infrared light is used as in the following embodiment, a near-infrared camera capable of receiving reflected light of such wavelengths may be used. The reflectance measuring device 14 may be any light source capable of measuring the reflectance (spectral reflectance) according to the wavelength of the light from the reflected light received by the photodetector 13, and may be a known device using a spectrophotometer. To measure reflectance more accurately, it is preferable to perform the light irradiation and the reception of reflected light by the photodetector 13 in an environment where the influence of external light is minimized. For example, it is preferable to place samples M1 to M3 in a darkroom for measurement. The light source 11, photodetector 13, and reflectance measuring instrument 14 can be commercially available. For example, the U4100 series spectrophotometer sold by Hitachi High-Tech Science Corporation is an integrated device that includes a wavelength-variable light source 11, a light irradiation unit 12, a photodetector 13, and a reflectance measuring instrument 14.

[0016] In the example of Figure 2, samples M1 to M3 for measuring the reflectance of each identified part are placed on three tables 15, with sample M1 consisting of only skin (with the hair removed), sample M2 consisting of only fat, and sample M3 consisting of only hair. Light is irradiated onto each of these samples M1 to M3, and the reflected light is received by the photodetector 13 to measure the reflectance of each sample M1 to M3. The measurement results are then calculated and processed in association with the type of meat and the parts to be distinguished, meaty and non-meaty, and the results are stored in the database DB as data for distinction.

[0017] [Database construction procedure] The procedure for constructing a database used for meat discrimination will be described below with reference to FIGS. As shown in FIG. 3, to start collecting data for meat discrimination, a data processing application pre-installed in the data processing device 10 is started (START). Next, a meat sample for which data is to be created is prepared, and sample information is input (Step S1). The sample information includes the type of meat (meaty part) (chicken in this example) and the parts of the non-meaty part to be identified (skin, fat, and hair in this example). Next, the initial and final wavelength values ​​of the light to be irradiated are input, and the division number n between them is input (step S2). For example, if the reflectance is to be measured every 1 nm in the wavelength range of 240 nm to 2600 nm, 240 nm is input as the initial value (k=1), 2600 nm as the final value (k=n), and the division number n is input as 2360. The range of wavelengths to be changed and the number of divisions n will vary depending on the type of meat, the type of parts to be distinguished in the meaty and non-meaty parts, the type of light to be used, etc., but the optimal range and number of divisions n will be selected based on experimentation and experience.

[0018] In this embodiment, light is irradiated onto samples of skin (with the hair removed, sample M1), fat (sample M2), and hair (sample M3) in that order, and the reflectance of each of samples M1, M2, and M3 is measured. After inputting information on the type of meat, samples M1 to M3, which are the non-meaty parts to be identified, the wavelength range, and the number of divisions n, the table 15 on which the first sample M1 is placed is placed under the light irradiator 12. Then, the table 15 on which the sample M1 is placed is positioned so that the light irradiated from the light irradiator 12 is reliably irradiated onto the sample M1 and the reflected light can be efficiently received by the photodetector 13 (step S3). In this case, it is advisable to perform a test irradiation of the light and position the sample M1 or the table 15 while checking on a monitor or the like the position at which the reflected light is efficiently received by the photodetector 13.

[0019] Once the positioning of the sample M1 is complete, light is irradiated onto a predetermined position on the sample M1 (step S4), the reflected light is received by the photodetector 13, and the reflectance of the sample M1 is measured by the reflectance measuring instrument 14 (step S5). It is preferable to measure the reflectance of the sample M1 multiple times. This may be done multiple times for one sample M1 by changing the light irradiation position, or by switching to another sample M1.

[0020] After irradiating light of one wavelength, receiving reflected light, and measuring the reflectance in this manner, the number of divisions k is counted and it is determined whether the specified number of divisions n has been reached (step S6). If the specified number of divisions n has not been reached, the number of divisions is set to k=k+1 and changed to the next wavelength (step S7). Thereafter, the irradiation of the sample M1 with light and the measurement of the reception and reflectance of the reflected light are repeated until the number of divisions k=n (steps S4 to S6). The measurement results for each wavelength are stored in the memory of the data processing device 10 in association with the wavelength.

[0021] When the division number k=n is reached (step S6), the data processing device 10 performs arithmetic processing on the reflectance obtained at each wavelength for sample M1. If multiple reflectances are obtained for each wavelength, the reflectance at that wavelength is determined, for example, by calculating the average or median. The reflectance for each wavelength for sample M1 can be visually displayed on a monitor or the like as a wavelength-reflectance graph, such as that shown in FIG. 4. In the graph in FIG. 4, the solid line indicates the relationship between wavelength and reflectance for sample M1 (skin).

[0022] After the irradiation of the first sample M1 with light and the measurement of its reflectance are completed, the process switches to the next sample M2 (step S9), and the process returns to step S3 to position sample M2 in the same way as sample M1. Thereafter, steps S4 to S8 are repeated. After the calculation of the measurement results in step S8 for all samples M1 to M3 is completed (step S9), two wavelengths at which the reflectances of the three samples M1, M2, and M3 are significantly different from each other are selected from the reflectance measurement results of each sample M1 to M3 (step S10). The greater the difference in reflectance at the same wavelength, the easier and more accurate it becomes to identify the target area by irradiating laser light of the same wavelength. In the example of Figure 4, the reflectances of the three samples M1, M2, and M3 are significantly different from one another within a range of about ±5 nm around 1205 nm and within a range of about ±40 nm around 1585 nm. Therefore, in step S10 of this embodiment, laser light with wavelengths of 1205 nm and 1585 nm is used to distinguish the skin (sample M1), fat (sample M2), and hair (sample M3) of the meat.

[0023] Furthermore, the data processing device 10 determines the range of reflectance for each wavelength, which serves as the basis for discrimination, from the distribution of multiple reflectances obtained by irradiating laser light multiple times for each of the two determined wavelengths, 1205 nm and 1585 nm (Figure 5). In the example of Figure 5, the distribution of the plot showing the reflectance of leather (sample M1) falls within the range of 23% to 28% at a wavelength of 1205 nm and the range of 10% to 15% at a wavelength of 1585 nm. If a tolerance of, for example, ±5% is set for each reflectance plot, a discrimination region I can be set that has a certain extent within the tolerance range of each reflectance. Alternatively, discrimination region I can be set using a pattern recognition technique such as a support vector machine (SVM). If the reflectance measured by irradiating two laser beams with wavelengths of 1205 nm and 1585 nm falls within this discrimination region I, the discriminated part can be discriminated as "skin," and the data processing device 10 can display the discriminated part on a monitor or the like using an easily identifiable color or pattern. The data processing device 10 associates the discrimination region I with the type of meat and the skin, and registers the association in the database DB (step S11).

[0024] Similarly, discrimination area II for fat (sample M2) and discrimination area III for hair (sample M3) are set from the plots of each reflectance using the same procedure as discrimination area I, and are associated with each other and registered in the database DB (step S11). This completes the construction of a database for meat discrimination for one type of meat (chicken in the above example) (END). In addition to domestic broilers and imported chicken from various countries, there are a wide variety of chicken breeds, including local chickens such as Hinai chicken and Nagoya Cochin, Silkie chickens, and Shamo chickens. Therefore, by repeating the same procedure as above for each type of chicken, it is possible to build a database that can more accurately identify meat from a wide variety of chicken breeds. Furthermore, by repeating the same procedures as above for other poultry meats such as duck, hybrid duck, and turkey, as well as meat from other animals such as beef, pork, lamb, and horse, it is possible to build a database that can identify meat according to the type of meat.

[0025] [Configuration of meat discrimination system] Next, an example of a meat discrimination system 1' that is installed in a meat processing plant or the like to discriminate meat will be described with reference to FIGS. FIG. 6 is a schematic diagram illustrating the configuration of a meat discrimination system 1′ installed in a meat processing plant or the like, FIG. 7 is a diagram showing an example of a meat discrimination result, and FIG. 8 is a flowchart illustrating the procedure for meat discrimination using the meat discrimination system. 6, the same members and parts as those of the data collecting device 1 in FIG. 1 are denoted by the same reference numerals, and detailed explanations thereof will be omitted.

[0026] When distinguishing between meaty and non-meaty parts of meat, it is necessary to identify the location of the meat where the distinguished parts (lean meat, skin, bone, fat, hair, etc.) are located. Therefore, in this embodiment, the meat (B1, B2, etc.) is irradiated with a sheet-like laser light, and the sheet-like laser light is scanned across the meat, thereby measuring the contours of the meat B1, B2, etc. in three dimensions using a known light cutting method, and by superimposing the reflectance measurement results based on the reflected light received by the photodetector 13 on the three-dimensional measurement results, it is possible to identify the position of the reflectance of each of the identified parts in the meat B1, B2, etc. It should be noted that although other methods than the light section method can be used for three-dimensional measurement of meat, such as a three-dimensional measurement method using multiple photodetectors, the light section method is advantageous in that it can simultaneously perform three-dimensional measurement of meat by irradiating it with laser light and measurement of reflected light.

[0027] The meat discrimination system of Figure 6 comprises a data processing device 10, a laser light source 11' that generates laser light of different wavelengths, a first light irradiator 121 that irradiates a sheet of laser light of one of the two wavelengths, a second light irradiator 122 that irradiates a sheet of laser light of the other wavelength, a first photodetector 131 that receives reflected light from an irradiated area with the laser light irradiated by the first light irradiator 121, and a second photodetector 132 that receives reflected light from an irradiated area with the laser light irradiated by the second light irradiator 122, and a reflectance measuring device 14 that measures reflectance from the reflected light received by the first and second photodetectors 131 and 132. The laser light source 11' may be separate sources that generate laser light of each wavelength, or it may be a single, tunable source. In order to make the discrimination more accurate, it is preferable that the irradiation of the laser light and the reception of the reflected light by the first and second photodetectors 131, 132 be carried out in an environment where the influence of external light is as small as possible. For example, it is preferable to provide a darkroom above the conveying process by the conveyor 16, and to irradiate the meat with the laser light and receive the reflected light within this darkroom. The data processing device 10 and the database DB are connected by a communication line N, and communication with the database DB becomes possible by starting up a meat discrimination application that has been installed in the data processing device 10 in advance.

[0028] The meats B1, B2, etc. are placed on a conveyor 16, and while being transported in the direction of the arrow, they are irradiated with sheet-like laser light from first light irradiator 121 and first light irradiator 122. First light irradiator 121 and first light detector 131 are arranged upstream in the transport direction, and second light irradiator 122 and second light detector 132 are arranged downstream in the same direction. As a result, two sheet-like laser light beams with different wavelengths are irradiated and scanned sequentially onto one piece of meat B1 (B2, etc.).

[0029] In order to further improve the accuracy of the discrimination results, it is preferable that the sheet-like laser light irradiated from the first light irradiating unit 121 and the second irradiating unit 122 is irradiated from the same direction onto one piece of meat B1 (B2...) and that the reflected light from each is received by the first and second photodetectors 131, 132 from the same direction, and it is preferable that the first light irradiating unit 121, the second irradiating unit 122 and the first and second photodetectors 131, 132 are each positioned and installed so that the direction of irradiation of the laser light and the direction of reception of the reflected light are the same.

[0030] Next, the procedure for meat discrimination using the meat discrimination system configured as described above will be described with reference to FIG. To start meat discrimination, a meat discrimination application pre-installed in the data processing device 10 is started (START), which enables communication with the database DB via the communication line N. Next, the type of meat to be identified is selected from the meats registered in the database DB (step S21). When the type of meat is selected, the two frequencies of laser light associated with the selected meat are identified from the data registered in the database DB (step S22).At the same time, the discrimination regions I, II, and III of reflectance required for discriminating the target part are also identified (step S23).

[0031] Once the above preparations are complete, click the Start Discrimination button on the data processing device 10 to output a command to start moving the conveyor 16 and to start irradiating laser light from the first light irradiating unit 121 and the second irradiating unit 122 (step S24). First, the first piece of meat B1 being conveyed on the conveyor 16 is irradiated with a sheet-like first laser light (laser light with a wavelength of 1205 nm) from the first light irradiator 121, and as it is scanned in the direction of the arrow, the reflected light is received by the photodetector 13. Next, the second light irradiator 122 irradiates the meat B1 with a sheet-like first laser light (laser light with a wavelength of 1585 nm), and as it is scanned in the direction of the arrow, the reflected light is received by the photodetector 13.

[0032] The reflectance of the reflected light received by the photodetector 13 is measured by the reflectance measuring device 14 (step S25), and the reflectance distribution of the meat B1 is calculated in the data processing device 10 from the three-dimensional measurement and the reflectance measurement results (step S26). Next, the data processing device 10 compares the reflectance for each wavelength with the threshold value registered in the database DB, and determines which of discrimination regions I, II, and III each reflectance falls into based on the wavelength, reflectance, and threshold value (step S27). The reflectance distribution thus sorted into discrimination regions I, II, and III is then allocated to the meat B1 obtained by three-dimensional measurement and displayed on a monitor or the like (step S28). An example of the display in step S8 is shown in Fig. 7. In Fig. 7, the bottom part in contact with the conveyor 16 is fat, which belongs to discrimination area II. The part above the fat is the skin, which belongs to discrimination area I, and on the surface of the skin is hair, which belongs to discrimination area III.

[0033] For example, meat that is deemed to require further processing, such as meat with thick hair attached to the skin, meat with an excessively thick or excessive fat layer attached to the skin, or meat with a large amount of fat protruding from the skin, can be detected by a monitor display or a warning light, and then removed from the conveyor 16 and sent to a processing step where the hair or excess fat can be removed. If the meat is determined to have thick hair or the ratio of fat to the skin exceeds a preset value, it is also possible to automatically transport the meat from the conveyor 16 to the processing step based on the detection result. The same applies to the other meats B2, B3, etc. that are conveyed on the conveyor 16 after the first meat B1.

[0034] Note that the fat hidden under the skin is not irradiated with laser light, and the presence of fat cannot be determined solely from the distribution of reflectance. Therefore, in this embodiment, the thickness of the skin is calculated from the height position of the skin surface using three-dimensional measurement, and if the skin thickness is greater than a certain value (e.g., greater than 2 mm), a monitor display or other means is used to alert the worker that fat may be present under the skin. Such meat may be transported from the conveyor 16 to a processing step.

[0035] [Other embodiments] Although the above description has been given using chicken as an example of a type of meat, the present invention can also be applied to other types of meat, such as beef and pork. In the following description, a case will be described in which meat portions and fat portions of beef and pork are distinguished from each other. Experiments were conducted on beef and pork under the following conditions. Wavelength of light used for data collection: 240~2600nm · Number of divisions n: 2360 (1 nm interval) Number of beef samples: 4 Measurements: 3 times each (total of 12 reflectance measurements)

[0036] [Distinguishing between meat and fat in beef] Data was collected using the same data collection equipment as for chicken meat described above. The results are shown in FIGS. FIG. 9 corresponds to FIG. 4 for chicken and shows the results of calculation processing of reflectance for beef, with graphs for each sample showing the relationship between wavelength and reflectance. FIG. 10 is a graph of the reflectance difference for each wavelength in the graph of FIG. 9. FIG. 11 corresponds to FIG. 5 and shows the distribution of reflectance for each sample at two wavelengths selected based on the calculation results of FIG. 9. From the graphs of FIGS. 9 and 10, 241 nm and 600 nm were selected as wavelengths at which there is a large difference in reflectance between the fleshy part and the fat. When a distribution diagram of the reflectance of each sample at these two wavelengths is created, it looks like Figure 11, with the meaty part and fat being distributed in two distinct regions. Therefore, similar to the embodiment for chicken, by setting a discrimination area for each of the meaty parts and fat and registering the threshold values ​​for setting these discrimination areas in the database, it becomes possible to distinguish between the meaty parts and fat in beef. If the discrimination method of the present invention is used for beef, for example, the distribution rate of fat in the meat part can be determined, and this can be used as a guide to determine the grade of beef based on the degree of marbling.

[0037] [Distinguishing between meat and fat in pork] Data was collected for pork using the same data collection device as for chicken, and the results are shown in Figures 12 to 14. FIG. 12 corresponds to FIG. 4 for chicken and shows the results of calculation processing of reflectance for pork, being a graph for each sample showing the relationship between wavelength and reflectance. FIG. 13 is a graph of the reflectance difference for each wavelength in the graph of FIG. 9. FIG. 14 corresponds to FIG. 5 for chicken and shows the distribution of reflectance for each sample at two wavelengths selected based on the calculation results of FIG. 12. From the graphs of FIGS. 12 and 13, 241 nm and 975 nm were selected as wavelengths at which there is a large difference in reflectance between the fleshy part and the fat. When a distribution diagram of the reflectance of each sample at these two wavelengths is created, it looks like Figure 14, with the meaty part and fat being distributed in two distinct regions. Therefore, similar to the embodiment for chicken, by setting a discrimination area for each of the meaty part and the fat, and registering the threshold value for setting this discrimination area in the database, it becomes possible to distinguish between the meaty part and the fat in pork.

[0038] Although the preferred embodiment of the present invention has been described above, the present invention is not limited to the above description. For example, although the skin, fat, and hair of non-meaty parts have been mentioned as the parts to be distinguished, the present invention can also be applied to distinguish other parts to be distinguished, such as bones and other foreign objects. Furthermore, while the above description has been about distinguishing non-meaty parts such as skin, bones, fat, and hair, it is also possible to distinguish differences in the quality of meat of the same type from the color of red meat, etc. Although laser light has been mentioned as the light to be irradiated for discrimination, light from light sources other than laser, such as natural light or illumination light, can also be used, and the wavelength can be changed, for example, by using a wavelength-variable filter. Furthermore, in the above explanation, the meat discrimination process was described as using light (laser light) of two wavelengths, but discrimination can also be performed using light of a single wavelength if there is a sufficiently large difference in reflectance between the wavelengths. In this case, the reflectance distribution will be as shown in Figure 15, and in this case too, discrimination areas I and II can be set as shown. [Industrial Applicability]

[0039] The present invention is not limited to chicken meat, but can also be applied to distinguish other meats such as beef, pork, lamb, and horse meat. [Brief explanation of the drawings]

[0040] [Figure 1] This is a photograph of the skin, fat, and hair, which are non-meaty parts of chicken, as an example of meat in this embodiment. [Figure 2] FIG. 1 is a schematic diagram of a data collection device for collecting data necessary for meat discrimination, creating the data, and storing it in a database. [Figure 3] 1 is a flow chart illustrating a procedure for collecting data. [Figure 4] 10 shows the results of calculation of reflectance, and is a graph showing the relationship between wavelength and reflectance for each sample. [Figure 5] FIG. 5 is a diagram showing the distribution of reflectance of each sample at two wavelengths selected based on the calculation results of FIG. 4. [Figure 6] 1 is a schematic diagram illustrating the configuration of a meat discrimination system installed in a meat processing plant or the like. [Figure 7] FIG. 10 is a diagram showing an example of a meat discrimination result. [Figure 8] 1 is a flowchart illustrating a procedure for meat discrimination by the meat discrimination system. [Figure 9] 1 shows the results of calculation of reflectance in the case of beef, and is a graph showing the relationship between wavelength and reflectance for each sample. [Figure 10] This is a graph of the reflectance difference for each wavelength in the graph of FIG. [Figure 11] FIG. 10 is a diagram showing the distribution of reflectance of each sample at two wavelengths selected based on the calculation results of FIG. 9. [Figure 12] 1 shows the results of calculation of reflectance in the case of pork, and is a graph showing the relationship between wavelength and reflectance for each sample. [Figure 13] This is a graph of the reflectance difference for each wavelength in the graph of FIG. [Figure 14] FIG. 13 is a diagram showing the distribution of reflectance of each sample at two wavelengths selected based on the calculation results of FIG. 12. [Figure 15] FIG. 10 is a diagram showing the distribution of reflectance of each sample when discrimination is performed using light of a single wavelength. [Explanation of symbols]

[0041] 1. Data collection equipment 1' Meat discrimination system 10 Data Processing Device 11 Light source 11' laser light source 12 Light irradiation unit 121 first light irradiation unit 122 Second light irradiation unit 13 Photodetector (light receiving means) 131 First photodetector (light receiving means) 132 Second photodetector (light receiving means) 14 Reflectance meter 15 tables 16 Conveyor B1,B2· Meat DB Database M1~M3 Samples N communication line

Claims

1. A meat discrimination method for discriminating between discriminated parts in a meat portion and / or a non-meat portion in a non-contact manner by irradiating light onto meat, The method comprises a preparation step of collecting data necessary for meat discrimination, and a discrimination step of discriminating between meat parts and / or non-meat parts to be discriminated in the meat, In the preparation step, a light source capable of irradiating the meat with light by changing the wavelength and a light receiving means for receiving reflected light from the meat; A step of changing the wavelength of light to be irradiated for each type of meat portion and each type of non-meat portion, and irradiating at least a portion of the meat; obtaining reflectance data for each of the identified portions in the meat portion and / or the non-meat portion of the meat from the reflectance of the meat portion and / or the reflectance of the non-meat portion for each wavelength obtained by irradiating the light; selecting at least one wavelength at which the difference in reflectance of each of the identified portions is greater than the others from the reflectance data and the wavelengths; a step of setting a discrimination region based on a distribution of reflectance for each of the discrimination regions at the selected wavelength for discrimination of each of the discrimination regions; storing the discrimination region for each of the discrimination target parts in a storage means; and In the meat discrimination process, a light source capable of irradiating the meat with light of at least one wavelength set in the preparation step, and a light receiving means for receiving reflected light from the meat; irradiating the entire meat with light of at least one of said wavelengths; a step of calculating a reflectance from the reflected light of the irradiated light, and comparing it with the discrimination area read out from the storage means to determine to which discrimination area of ​​the discrimination portion the reflectance belongs; a step of identifying the position of reflected light belonging to any of the discrimination regions and displaying the distribution of each of the discrimination regions in the meat; A meat discrimination method comprising the steps of:

2. 2. The meat discrimination method according to claim 1, wherein the discrimination region is set based on reflectance data for a plurality of different wavelengths.

3. 3. The meat discrimination method according to claim 1, wherein the meat is measured in three dimensions, and the discriminated portions of the meaty portion and / or the non-meaty portion are discriminated from the results of the three-dimensional measurement and the reflectance.

4. 3. The meat discrimination method according to claim 1 or 2, wherein the light of the selected wavelength is irradiated onto the entire meat while scanning it, and the position of the scanned area is identified from the reflection angle of the reflected light, thereby simultaneously performing the three-dimensional measurement and discriminating between the meaty parts and the non-meaty parts based on the reflectance.

5. 5. The meat discrimination method according to claim 3 or 4, wherein the thickness of the non-meaty parts is determined from the results of the three-dimensional measurement, and when the thickness of the non-meaty parts obtained and the type of the non-meaty parts exceed a predetermined threshold, it is determined that different types of non-meaty parts are layered together.

6. 3. The method for distinguishing meat according to claim 1, wherein the meat portion is chicken meat, and the non-meat portion is fat, skin, or hair.

7. 7. The meat discrimination method according to claim 6, wherein the discrimination region is set based on reflectance at wavelengths of 1200±50 nm and 1550±50 nm.

8. A method for constructing a database for meat discrimination in which a discrimination target part in a meat part and / or a non-meat part is discriminated in a non-contact manner by irradiating meat with light, comprising: a light source capable of irradiating the meat with light by changing the wavelength and a light receiving means for receiving reflected light from the meat; A step of changing the wavelength of light to be irradiated for each type of meat portion and each type of non-meat portion, and irradiating at least a portion of the meat; obtaining reflectance data for each of the identified portions of the meat from the reflectances of the meat portions and the non-meat portions for each wavelength obtained by irradiating the light; selecting at least one wavelength at which the difference in reflectance between the identified portions is greater than the other wavelengths from the reflectance data of the fleshy portion or the non-fleshy portion and the wavelengths; setting a discrimination region for each of the discrimination regions from a distribution of reflectance for each of the discrimination regions at the selected wavelength; storing the discrimination region for each of the discrimination target parts in a storage means; A database construction method comprising:

9. 9. The database construction method according to claim 8, wherein the discrimination region is set from reflectance data for a plurality of different wavelengths.

10. a step of storing the thickness of the non-meaty portion as thickness data from the results obtained by the three-dimensional measurement of the meat; determining a thickness as a threshold value to be used for determining that a non-discrimination portion is layered on a discrimination portion of the non-discrimination portion based on the thickness data and the type of the non-discrimination portion, and storing the determined thickness; 10. The database construction method according to claim 8, further comprising:

11. A meat discrimination system that comprises a database constructed by the method according to any one of claims 8 to 10, and discriminates between meaty parts and non-meaty parts in a non-contact manner by irradiating meat with light, a light source that irradiates light of one or more wavelengths predetermined according to the type of meat; an irradiation means for irradiating at least a portion of the meat with the light; a light receiving means for receiving reflected light of the light; a reflectance measuring means for determining a reflectance from the reflected light received by the light receiving means; a calculation means connected to the database, which reads out a discrimination area for each of the discriminated parts according to the type of meat, determines which of the discriminated parts the discriminant area the measured light reflectance belongs to, and specifies where in the meat the reflectance is located from the three-dimensional measurement results of the meat; A meat discrimination system comprising:

Images