Device and method for identifying male and female information of ducklings based on spectral analysis technology

Abstract

The invention discloses a device and a method for identifying male and female information of ducklings based on a spectral analysis technology, and relates to a poultry male and female identification technology in the field of agricultural machinery. The device is characterized in that a duckling is placed on a sample collection platform (3), and an output end probe of a Y-shaped optical fiber (5) is aligned with a cloaca of the duckling; the spectrograph (6) and the halogen tungsten lamp light source (7) are respectively fixed at the lower part of the inner cavity of the box body (1) through a spectrograph fixing frame (9) and a light source fixing frame (10); the input end of the Y-shaped optical fiber (5) is respectively connected with a spectrograph (6) and a halogen tungsten lamp light source (7); and the spectrograph (6) is connected with the computer (8). The method comprises the following steps: (1) sample preparation; (2) spectrum acquisition; (3) data processing; (4) sample analysis; and (5) an identification result display interface. The method has the advantages of high precision, fast speed and no harm to ducklings; (2) an identification result display interface is visual and clear;

Description

Technical Field

[0001] The invention relates to a technology for identifying the sex of poultry in the field of agricultural machinery, and in particular to a device and a method for identifying the sex of ducklings based on spectral analysis technology. Background Art

[0002] In the process of large-scale duckling breeding, there are differences in growth and development and dietary requirements between male and female ducklings. Female ducklings are usually raised as commercial laying ducks, and the core goal of their growth and development is egg-laying performance. In order to ensure high egg production and good egg quality, female ducklings need to be provided with appropriate nutrients such as protein and vitamins. Male ducklings are usually raised as meat ducks, and the focus of their growth and development is muscle growth and weight gain. In order to improve the lean meat rate and meat taste of meat ducks, male ducklings need to be provided with appropriate nutrients such as energy and amino acids. When male and female ducklings are mixed, due to their different nutritional needs, if the same feed formula is used, it may cause nutritional imbalance, poor growth and development, feed waste and increased breeding costs, affecting breeding benefits. Therefore, it is necessary to adjust the feeding management and feed formula in a targeted manner to meet their respective needs.

[0003] In order to reduce feed waste and breeding costs and increase breeding benefits, it is of great significance to study the technology of gender identification of ducklings. At present, the traditional manual identification methods are mainly used for gender identification of ducklings, including: appearance identification method, syrinx identification method and anus identification method. The difference in the appearance of ducklings is difficult to distinguish with the naked eye, and the accuracy rate is not high; although the syrinx identification method can also be distinguished, there are problems of eye fatigue and reliance on experience; the anus identification method is the most used, with a high accuracy rate, but it requires long-term and rich operating experience and has high labor intensity. In order to accelerate the development of modern duck farming, adapt to large-scale farming, reduce the labor intensity of farmers, and improve the level of intelligent equipment in duck farming, it is necessary to develop a non-destructive, easy-to-operate and efficient technology for gender identification of ducklings.

[0004] Near-infrared spectroscopy detection technology uses visible / near-infrared light (400-2526nm) to analyze the chemical composition and structural information of the object to be tested by absorbing the vibration frequency doubling and combined frequency of hydrogen-containing groups (-CH, -OH, -NH, etc.). Since male ducklings have a reproductive protrusion of about 2-4mm (2-4mm soft cartilage), female ducklings do not. The reproductive protrusion is mainly composed of fibrous lymphoid bodies, connective tissue, lymphatic spaces, microvessels and smooth muscles. The components of these tissues mainly include water, protein, lipids and cytochrome. Therefore, the near-infrared spectra of the physiological structures of male and female ducklings will be different, and it is feasible to use near-infrared spectroscopy detection technology to identify the male and female information of ducklings. By scanning the near-infrared spectrum of the duckling cloaca, the spectral data of the male and female information of the duckling can be obtained. The gender of ducklings is manually identified by experienced duckling sex identifiers, and the near-infrared spectral data of the ducklings is associated with the gender information of the ducklings. The gender of the ducklings can be identified through spectral preprocessing, characteristic wavelength extraction and construction of a classification model. Summary of the invention

[0005] The purpose of the present invention is to provide a device and method for identifying the gender of ducklings based on spectral analysis technology, in order to address the problems of high skill and labor intensity in traditional manual identification of the gender of ducklings, which can achieve high-accuracy, fast, non-destructive and easy-to-operate identification of the gender of ducklings.

[0006] The objective of the present invention is achieved through the following technical solutions:

[0007] 1. Duckling gender identification device based on spectral analysis technology

[0008] The device includes a box, light-blocking cotton, a sample collection platform, an optical fiber fixing fixture, a Y-type optical fiber, a spectrometer, a halogen tungsten light source, a computer, a spectrometer fixing frame and a light source fixing frame;

[0009] Its location and connection relationship are:

[0010] The top surface of the box is provided with light-blocking cotton and a sample collection platform connected up and down;

[0011] The upper, middle and lower parts of the inner cavity of the box are provided with an optical fiber fixing fixture, a Y-type optical fiber, a spectrometer, a halogen tungsten light source, a spectrometer fixing frame and a light source fixing frame in sequence;

[0012] A computer is arranged outside the box;

[0013] Place the duckling on the sample collection platform, and aim the output end probe of the Y-type optical fiber at the duckling's cloaca;

[0014] The spectrometer and the halogen tungsten lamp light source are fixed to the lower part of the inner cavity of the box through a spectrometer fixing frame and a light source fixing frame respectively;

[0015] The input ends of the Y-type optical fiber are connected to the spectrometer and the halogen tungsten light source respectively;

[0016] Connect the spectrometer to the computer.

[0017] 2. Method for identifying the sex of ducklings based on spectral analysis technology

[0018] This method includes five parts: sample preparation, spectrum acquisition, data processing, sample analysis and identification result display. Specifically, the detection process includes:

[0019] ① Sample preparation: The test samples were freshly hatched ducklings of the Guiliu M18 breed from a local duck farm. Experienced duckling gender identification experts identified 25 female ducklings and 25 male ducklings.

[0020] ② Spectral acquisition: Before collecting the spectral data of the male and female information of the ducklings, initialize the spectrometer 6 first, turn on the halogen tungsten light source 7 to preheat for 15 minutes, and open the spectral acquisition software supporting the spectrometer 6, set the acquisition parameters as follows: integration time is 50ms, average scanning number is 10, smoothness is 1, and set the save location of the spectral data; then align the duckling cloaca with the probe of the Y-type optical fiber 5 and press down, so that the probe and the light-shielding cotton 2 are close to the duckling's cloaca; finally, click the download button in the acquisition software interface to save the collected raw spectral data of the male and female information of the ducklings to the specified location, and the spectral data is saved in the format of a txt file;

[0021] ③Data processing: Based on Python language, a program was designed in Pycharm software to perform Savitzky-Golay smoothing preprocessing on the collected raw spectral data, and the spectral data corresponding to the characteristic wavelength extracted by the UVE-SPA algorithm were selected;

[0022] ④ Sample analysis: Load the mobile network V2 (MobileNetV2) with the optimal weight, import the processed spectral data into MobileNetV2 for analysis, and finally obtain the identification result; after the classification is completed, the spectral data file will be automatically deleted to prepare for the spectral data collection of the next sample;

[0023] ⑤Identification result display interface: The identification result display interface was designed using the PyQt5 library in the Pycharm software. The gender information and analysis time of the ducklings analyzed during the sample analysis process were displayed in real time through the identification result display interface.

[0024] Compared with the prior art, the present invention has the following advantages and positive effects:

[0025] ①High precision, fast speed, no harm to ducklings;

[0026] ②The identification result display interface is intuitive and clear;

[0027] ③ After the designed data processing program pre-processes the spectral data, it extracts characteristic bands and removes a large number of repeated and redundant bands that are irrelevant to the gender information of ducklings. Only characteristic bands are used for prediction, which saves prediction time, improves the identification speed, and reduces the impact of environmental changes on prediction accuracy.

[0028] ④The software design includes functions such as data processing, sample analysis and identification result display. Compared with the traditional manual identification method, it lowers the usage threshold. Even users who do not know manual identification skills can operate it with less labor intensity. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a structural block diagram of the device;

[0030] Figure 2 It is a structural schematic diagram of the device;

[0031] Figure 3 is a schematic diagram of the structure of an optical fiber fixing fixture;

[0032] Figure 4 It is a schematic diagram of the structure of the spectrometer fixing frame;

[0033] Figure 5 It is a structural schematic diagram of a light source fixing frame;

[0034] Figure 6 This is the identification flow chart of this method.

[0035] In the figure:

[0036] 0—duckling;

[0037] 1—Box;

[0038] 2—light-blocking cotton;

[0039] 3—Sample collection platform;

[0040] 4—optical fiber fixing fixture, 4-1—fixing column, 4-2—fixing ring, 4-3—double-sided spring, 4-4—fixing cylinder;

[0041] 5—Y-type optical fiber;

[0042] 6—Spectrometer;

[0043] 7—Halogen tungsten light source;

[0044] 8—Computer;

[0045] 9—Spectrometer fixing bracket, 9-1—Spectrometer card slot, 9-2—Spectrometer C-type fixing bracket;

[0046] 10—light source fixing frame, 10-1—light source slot, 10-2—light source C-type fixing frame. DETAILED DESCRIPTION

[0047] The following is a detailed description with reference to the accompanying drawings and embodiments.

[0048] 1. Device

[0049] 1. Overall

[0050] like Figure 1 , 2 The device includes a box 1, light-blocking cotton 2, a sample collection platform 3, an optical fiber fixing fixture 4, a Y-type optical fiber 5, a spectrometer 6, a halogen tungsten light source 7, a computer 8, a spectrometer fixing frame 9 and a light source fixing frame 10;

[0051] Its location and connection relationship are:

[0052] A light-blocking cotton 2 and a sample collection platform 3 connected up and down are provided on the top surface of the box body 1;

[0053] The upper, middle and lower parts of the inner cavity of the box body 1 are provided with an optical fiber fixing fixture 4, a Y-type optical fiber 5, a spectrometer 6, a halogen tungsten light source 7, a spectrometer fixing frame 9 and a light source fixing frame 10 in sequence;

[0054] A computer 8 is provided outside the box 1;

[0055] Place the duckling on the sample collection platform 3, and aim the output end probe of the Y-type optical fiber 5 at the duckling's cloaca;

[0056] The spectrometer 6 and the halogen tungsten light source 7 are fixed to the lower part of the inner cavity of the box body 1 through the spectrometer fixing frame 9 and the light source fixing frame 10 respectively;

[0057] The input ends of the Y-type optical fiber 5 are connected to a spectrometer 6 and a halogen tungsten lamp light source 7 respectively;

[0058] The output end of the spectrometer 6 is connected to a computer 8 .

[0059] 2. Functional components

[0060] 01) Cabinet 1

[0061] The box 1 is a support body for the functional components, and has a length×width×height=450mm×450mm×520mm. It not only ensures the transmission performance of the Y-type optical fiber 5, but also makes the entire box 1 as small as possible for easy transportation.

[0062] 02) Light blocking cotton 2

[0063] The light-blocking cotton 2 is closely attached to the cloaca of the duckling to ensure the safety of the duckling's cloaca.

[0064] 03) Sample collection platform 3

[0065] The sample collection platform 3 is a flat plate for placing the duckling 0.

[0066] 04) Fiber optic fixture 4

[0067] It is made of polylactic acid material and is manufactured using a 3D printer; it is used to fix the Y-shaped optical fiber 5.

[0068] like Figure 3 , the optical fiber fixing fixture 4 includes a fixing column 4-1, a fixing ring 4-2, a double-sided spring 4-3 and a fixing cylinder 4-4;

[0069] The connection relationship is:

[0070] With the Y-type optical fiber 5 as the central axis, the fixing column 4-1, the fixing ring 4-2, the double-sided spring 4-3 and the fixing tube 4-4 are connected in sequence; the top surface of the fixing column 4-1 is fitted and fixed under the top surface of the inner cavity of the box body 1.

[0071] A. Fixed column 4-1

[0072] The fixed column 4-1 is a hollow T-shaped cylinder.

[0073] B. Fixed ring 4-2

[0074] The fixing ring 4 - 2 is a circular ring with an adjustable inner diameter, and the inner diameter of the circular ring can be adjusted to match the Y-shaped optical fiber 5 .

[0075] C. Double-sided spring 4-3

[0076] The double-sided spring 4-3 is a short spring.

[0077] D. Fixed tube 4-4

[0078] The fixed cylinder 4-4 is a cylinder.

[0079] 05)Y-type optical fiber 5

[0080] The Y-type optical fiber 5 adopts FIB-IR600-Y7TA-1.0 of Winyi Optoelectronics. The probe diameter of the Y-type optical fiber 5 is 6.35mm, the optical fiber length is 1m, the number of cores is 7 and the connector is SMA905; its input end is used to connect the spectrometer 6 and the halogen tungsten lamp light source 7, and its output end is used to connect the probe.

[0081] 06) Spectrometer 6

[0082] Spectrometer 6 uses the NIRQUEST+2.5 high-performance spectrometer from Ocean Optics, USA, with the following parameters: linear array InGaAs detector; the detector band range is 898-2495nm; dynamic pixels are 512, the pixel size is 25um×250um, and the optical resolution is 6.3nm.

[0083] Collect a spectrum of the cloaca of a duckling.

[0084] 07) Halogen tungsten light source 7

[0085] The halogen tungsten lamp light source 7 adopts the Wenyi Optoelectronics HL-1000 halogen tungsten lamp light source, and its parameters are: wavelength range 360-2500nm; color temperature 3000K; power consumption 20W; working life 2000H.

[0086] Provide a light source.

[0087] 08) Computer 8

[0088] The interface of the computer 8 is USB 2.0.

[0089] Perform data calculations and display results.

[0090] 09)Spectrometer fixing bracket 9

[0091] It is made of polylactic acid material and is made by a 3D printer; it is used to fix the spectrometer at the bottom of the inner cavity of the box 1.

[0092] like Figure 4 The spectrometer fixing frame 9 includes spectrometer card slots 9-1 symmetrically placed on both sides and a spectrometer C-type fixing frame 9-2 placed in the middle.

[0093] 10) Light source fixing frame 10

[0094] The polylactic acid material is used and made by a 3D printer. The halogen tungsten lamp light source 7 is fixed at the bottom of the inner cavity of the box body 1.

[0095] like Figure 5 The light source fixing frame 10 includes light source card slots 10 - 1 symmetrically placed on both sides and a light source C-shaped fixing frame 10 - 2 placed in the middle.

[0096] 3. Working mechanism:

[0097] The device includes a sample collection platform 3, on which a probe of a Y-type optical fiber 5 is arranged, which is fixed on the sample collection platform 3 by an optical fiber fixing fixture 4, and the probe is 20 mm long, and a layer of 7 mm thick light-blocking cotton is arranged on the sample collection platform 3, which surrounds the probe of the Y-type optical fiber 5, and is used to prevent light leakage from affecting the collected spectral data and protect the duckling cloaca when collecting the spectral data of the male and female information of the duckling. The probe of the Y-type optical fiber 5 is connected to a halogen tungsten lamp light source 7 through the illumination optical fiber of the Y-type optical fiber 5 and is connected to a spectrometer 6 through the detection optical fiber of the Y-type optical fiber 5. The halogen tungsten lamp light source adopts an adjustable halogen tungsten lamp capable of emitting a 360-2500nm spectrum line, and the spectrometer 6 can obtain a near-infrared spectrum with an effective range of 898-2495nm. The spectrometer 6 is connected to a computer 8 through a USB, and the spectrum collection software and identification program are operated on the computer 8 to realize the identification of the male and female information of the duckling.

[0098] In the above identification device, the box 1 contains a spectrometer 6, a halogen tungsten light source 7, a Y-type optical fiber 5 and an optical fiber fixing fixture 4. The spectrometer 6 and the halogen tungsten light source 7 are fixed to the bottom of the box by a spectrometer fixing frame 9 and a light source fixing frame 10.

[0099] Both fixing frames adopt the same structure, i.e., a card slot and a C-type fixing frame, which are installed at the bottom of the inner cavity of the box body 1 by means of bolts and nuts. The optical fiber fixing fixture 4 is fixed above the inner cavity of the box body 1, i.e., the back of the sample collection platform 3, by means of a fixing column 4-1 and bolts and nuts. The Y-type optical fiber 5 is fixed on the double-sided spring 4-3 by means of an optical fiber fixing ring 4-2 and bolts and nuts. The double-sided spring 4-3 is provided with a traceless nano-glue and can be fixed in a fixing cylinder 4-4. The fixing cylinder 4-4 is installed on the fixing column 4-1 by means of bolts and nuts, and there is an 8mm hole at the bottom of the fixing cylinder 4-4 for the Y-type optical fiber 5 to pass through. The optical fiber fixing fixture 4 can not only fix the Y-type optical fiber 5, but also, when the duckling's cloaca is placed on the probe of the Y-type optical fiber 5 to collect spectral data, the double-sided spring 4-3 is in a compressed state, which will give a reverse pressure, so that the probe of the Y-type optical fiber 5 is close to the duckling's cloaca, thereby preventing the influence of light leakage on the collected spectral data. To ensure the transmission performance of the Y-type optical fiber 5, its bending radius cannot be less than 9.9 cm, so the size of the box 1 is designed to be length × width × height = 450 mm × 450 mm × 520 mm, which not only ensures the transmission performance of the Y-type optical fiber 5, but also makes the entire box 1 as small as possible for easy transportation.

[0100] 2. Methods

[0101] The identification process of the above identification device is summarized as follows:

[0102] After the spectrometer 6 is initialized, turn on the spectrometer and set the parameters on the spectrum acquisition software, respectively setting the integration time, smoothness and average scan number;

[0103] Then, set the storage path of the spectral data file, that is, the folder path of the classification model input data;

[0104] Finally, click the Download button to save the collected spectral data (txt format file) to the specified input folder;

[0105] Once the txt file is saved to the input folder, the classification program will first perform Savitzky-Golay smoothing preprocessing on the raw spectral data in the file and select the spectral data corresponding to the characteristic wavelengths extracted by the UVE-SPA algorithm;

[0106] Then the spectral data is imported into MobileNetV2 loaded with the optimal weights for analysis, the identification result is obtained and the file is deleted;

[0107] Finally, the identification result display interface designed using the PyQt5 library in the Pycharm software displays the identification results in real time.

[0108] like Figure 6 , the identification process is as follows:

[0109] ① Near infrared spectrum collection of ducklings-610

[0110] Ⅰ. Aim at the collection site-611;

[0111] Ⅱ. Open the spectrum acquisition software and set the acquisition parameters to -612;

[0112] III. Click Download to collect spectral data-613 and jump to step ii;

[0113] or ②Data material and sample analysis-620

[0114] ⅰ. Open the authentication program-621;

[0115] ii. Savitzky-Golay convolution smoothing - 622;

[0116] ⅲ, spectral data at characteristic wavelengths extracted by UVE-SPA algorithm - 623;

[0117] iv. Import the classification model for sample prediction-624, jump to step A;

[0118] Or ③ ​​The identification result shows -630

[0119] A. PyQt5 library design display interface-631.

[0120] The above embodiments are only for illustrating the technical idea of ​​the present invention, and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the present invention; any technology not involved in the present invention can be realized by existing technologies.

Claims

1. A device for identifying the gender of ducklings based on spectral analysis technology, characterized in that: It comprises a box (1), light-blocking cotton (2), a sample collection platform (3), an optical fiber fixing fixture (4), a Y-shaped optical fiber (5), a spectrometer (6), a halogen tungsten light source (7), a computer (8), a spectrometer fixing frame (9) and a light source fixing frame (10); Its location and connection relationship are: A light-blocking cotton (2) and a sample collection platform (3) are provided on the top surface of the box body (1); An optical fiber fixing fixture (4), a Y-shaped optical fiber (5), a spectrometer (6), a halogen tungsten light source (7), a spectrometer fixing frame (9), and a light source fixing frame (10) are sequentially arranged at the upper, middle, and lower parts of the inner cavity of the box body (1); A computer (8) is arranged outside the box (1); The duckling (0) is placed on the sample collection platform (3), and the output end probe of the Y-type optical fiber (5) is aimed at the duckling's cloaca; The spectrometer (6) and the halogen tungsten lamp light source (7) are fixed to the lower part of the inner cavity of the box body (1) via a spectrometer fixing frame (9) and a light source fixing frame (10) respectively; The input ends of the Y-shaped optical fiber (5) are respectively connected to a spectrometer (6) and a halogen tungsten lamp light source (7); The output end of the spectrometer (6) is connected to a computer (8) to obtain spectrum data.

2. The device for identifying the gender of ducklings based on spectral analysis technology according to claim 1, characterized in that: The Y-type optical fiber (5) adopts FIB-IR600-Y7TA-1.0 of Wenyi Optoelectronics. The probe diameter of the Y-type optical fiber 5 is 6.35mm, the optical fiber length is 1m, the number of roots is 7 and the connector is SMA905; its input end is used to connect the spectrometer 6 and the halogen tungsten lamp light source 7, and its output end is used to connect the probe.

3. The device for identifying the gender of ducklings based on spectral analysis technology according to claim 1, characterized in that: The spectrometer (6) uses the NIRQUEST+2.5 high-performance spectrometer of the American Ocean Optics Company, and its parameters are: linear array InGaAs detector; the detector band range is 898-2495nm; dynamic pixels are 512, the pixel size is 25um×250um, and the optical resolution is 6.3nm, which collects the spectrum of the cloaca of the ducklings.

4. The device for identifying the gender of ducklings based on spectral analysis technology according to claim 1, characterized in that: The halogen tungsten lamp light source (7) adopts the Wenyi Optoelectronics HL-1000 halogen tungsten lamp light source, and its parameters are: wavelength range 360-2500nm; color temperature 3000K; power consumption 20W; working life 2000H.

5. The identification method based on the device of claim 1, 2, 3 or 4, characterized in that: ① Sample preparation: The test samples were freshly hatched ducklings of the Guiliu M18 breed from a local duck farm. Experienced duckling gender identification experts identified 25 female ducklings and 25 male ducklings. ② Spectral acquisition: Before collecting the spectral data of the male and female information of the ducklings, initialize the spectrometer 6 first, turn on the halogen tungsten light source 7 to preheat for 15 minutes, and open the spectral acquisition software supporting the spectrometer 6, set the acquisition parameters as follows: integration time is 50ms, average scanning number is 10, smoothness is 1, and set the save location of the spectral data; then align the duckling cloaca with the probe of the Y-type optical fiber 5 and press down, so that the probe and the light-shielding cotton 2 are close to the duckling's cloaca; finally, click the download button in the acquisition software interface to save the collected raw spectral data of the male and female information of the ducklings to the specified location, and the spectral data is saved in the format of a txt file; ③Data processing: Based on Python language, a program was designed in Pycharm software to perform Savitzky-Golay smoothing preprocessing on the collected raw spectral data, and the spectral data corresponding to the characteristic wavelength extracted by the UVE-SPA algorithm were selected; ④ Sample analysis: Load the optimal weighted mobile network V2 (MobileNetV2), import the processed spectral data into MobileNetV2 for analysis, and finally obtain the identification result; After the classification is completed, the spectral data file will be automatically deleted to prepare for the spectral data collection of the next sample; ⑤Identification result display interface: The identification result display interface was designed using the PyQt5 library in the Pycharm software. The gender information and analysis time of the ducklings analyzed during the sample analysis process were displayed in real time through the identification result display interface.

6. The identification method according to claim 5, characterized in that: The identification process is as follows: ① Near infrared spectrum collection of ducklings (610) Ⅰ. Aligning the collection site (611); II. Open the spectrum acquisition software and set the acquisition parameters (612); III. Click download to collect spectral data (613), and jump to step ii; or ②Data materials and sample analysis (620) i. Open the authentication program (621); ii. Savitzky-Golay convolution smoothing (622); iii. Spectral data at characteristic wavelengths extracted by UVE-SPA algorithm (623); iv. Import the classification model to perform sample prediction (624), and jump to step A; or ③ identification result display (630) A. PyQt5 library designs the display interface (631).

Images