An embryo screening method and device based on millimeter-wave radar

Abstract

This application discloses a method and apparatus for embryo screening based on millimeter-wave radar. The method includes: irradiating an embryo egg with a frequency-modulated continuous wave emitted by a millimeter-wave radar; receiving the echo signal and mixing it with the emitted signal to obtain an intermediate frequency (IF) signal; transmitting the IF signal to a computer signal processing terminal via a signal acquisition device; preprocessing the radar signal, including performing circle fitting using a correlation-optimized least squares method to eliminate DC offset, and suppressing static clutter components using a phasor mean cancellation algorithm; constructing a range-angle map and locating multiple embryo eggs using constant false alarm rate (CFAR); extracting the phase information of the target from the preprocessed signal to obtain the embryo's heartbeat frequency, and combining this with the embryo type and incubation duration to determine the embryo's incubation status in real time. This invention allows for accurate, real-time, and non-contact detection of the internal heartbeat frequency of an embryo egg, determining the incubation status of the embryo under test, effectively improving detection efficiency and reducing resource consumption.

Description

Technical Field

[0001] This invention relates to the field of embryo hatching technology, specifically to an embryo screening method and device based on millimeter-wave radar. Background Technology

[0002] With the rapid development of the global poultry farming industry, my country ranks first in the world in poultry egg production, making the selection of eggs for hatching increasingly important. Before eggs are placed in incubators, embryos need to be identified and classified to remove dead embryos, infertile eggs, and rotten eggs, while live embryos are left to hatch. This avoids cross-contamination, effectively reduces resource consumption, and improves hatching efficiency and egg quality.

[0003] Early hatcheries used manual candling to inspect eggs. This involved manually shining a cannon on the larger end of the egg to visually observe its internal development, such as the air cell and blood vessel morphology, and then selecting infertile, dead, or rotten eggs. This method was not only extremely inefficient and lacked specific quantifiable standards, but its accuracy also decreased over time. If dead or rotten eggs were mistakenly accepted as qualified eggs and continued to be incubated, it not only wasted incubator space but also contaminated other qualified eggs, significantly reducing incubation efficiency.

[0004] Currently, the main detection methods used in domestic and international research on poultry egg embryo culture technology include four methods: ACG (acoustocardiogram), BCG (ballistocardiogram), ECG (electrocardiogram), and ICG (impedance-cardiogram). ECG and ICG methods involve drilling holes in the eggshell and inserting electrodes to measure the characteristic signals of the living poultry embryo. Because the electrodes are inserted relatively deeply into the embryo, this method can damage the embryo's physiological structure. After the test, the tested poultry egg embryo stops developing, making it an invasive measurement method that cannot meet the production needs of the hatching industry. ACG and BCG methods, on the other hand, are non-invasive measurement methods. The ACG method measures the sound signals emitted by a living poultry egg embryo during incubation by placing the egg directly in a sealed container for measurement. The BCG method measures the fetal movement signals of a living poultry embryo during incubation by attaching a magnet to the eggshell and using a fixed electromagnetic coil to convert the tiny movements of the magnet into an induced current for measurement. The signals extracted by these methods are extremely weak and require very strict testing environments.

[0005] The scale of the hatching industry is constantly expanding, and existing poultry egg screening methods are insufficient to meet the requirements of large-scale operations. Therefore, there is an urgent need to develop an automated embryo detection device to replace manual egg candling for screening, in order to meet the needs of large-scale poultry hatcheries and improve the accuracy and efficiency of embryo egg screening. Summary of the Invention

[0006] This application discloses an embryo screening method and device based on millimeter-wave radar, which enables real-time and rapid detection of embryo development at any location in the early stage of poultry egg incubation, timely screening of infertile eggs, dead embryos and other embryonic eggs, improving detection efficiency and reducing resource consumption.

[0007] The technical solution of the present invention is as follows:

[0008] In the first aspect, an embryo egg screening and detection device is provided, including an embryo incubation device, a millimeter-wave radar motion detection device, a multi-point signal acquisition device, and a computer signal processing terminal.

[0009] The embryo incubation device consists of an incubation box and an egg-fixing container, which is used to hold multiple embryo eggs and provide a suitable incubation environment;

[0010] The millimeter-wave radar motion detection device consists of a millimeter-wave radar sensor, a robotic arm support, and a sliding guide rail. The millimeter-wave radar sensor is used to send a frequency-tunable continuous wave (FMCW) into the embryonic egg and receive the echo signal, which is then mixed with the transmitted signal. The robotic arm support is used to adjust the millimeter-wave radar sensor so that it is directly facing the center of the embryonic egg incubation device. The sliding guide rail is used to move the millimeter-wave radar along the sliding guide rail, thereby enabling the detection and screening of all embryonic eggs inside the embryonic incubation device along the movement path.

[0011] The multi-point signal acquisition device is used to receive data sent by the radar and send it to the computer signal processing terminal.

[0012] The computer signal processing terminal is used to receive radar data, perform signal processing, separate the embryonic vital signs information and interference signals in the data, amplify and extract the heartbeat signal of the internal embryo.

[0013] Secondly, a detection method for embryo screening is provided, comprising the following steps:

[0014] The millimeter-wave radar sensor is controlled to send a frequency-tunable continuous wave (FMCW) into the embryonic egg, and the echo signal is received and mixed with the transmitted signal to obtain an intermediate frequency signal.

[0015] The mixed intermediate frequency signal data is sent to a computer signal processing terminal for signal processing via a signal acquisition device.

[0016] The radar signal is preprocessed, including: performing range-FFT and angle-FFT on each chirp of the radar signal to convert the time-domain data into frequency-domain data; using the correlation-optimized least squares method to perform circle fitting on the radar data to eliminate DC offset; using the phasor mean cancellation algorithm to suppress static clutter components in the radar data; constructing a range-angle map to perform two-dimensional localization of multiple embryos in the detection area; and using the constant false alarm rate (CFAR) algorithm to determine the number and location of embryos within the detection area.

[0017] Phase information is extracted from the preprocessed frequency domain signal, including: extracting the phase information at the maximum position after distance-FFT by phase arctangent, and unwinding the phase to ensure the continuity of phase change; calculating the phase difference of the extracted phase to enhance the embryo's heartbeat signal.

[0018] Furthermore, an 8th-order IIR bandpass filter is used to extract the embryonic heartbeat signal, and a time-domain peak-finding algorithm is used to find the peak of the heartbeat signal. The frequency of the peak is then calculated to obtain the embryonic heartbeat frequency.

[0019] Furthermore, by collecting and calculating the number of heartbeats per minute of the embryos, and based on the incubation time of the tested embryo eggs and the type of embryo being tested, the survival status of the tested embryos is determined.

[0020] This invention provides a non-contact detection method for embryo screening based on millimeter-wave radar. It can effectively detect the heartbeat information inside the embryo egg, thereby determining whether the embryo is viable. This solves the problems of interference and potential damage to the embryo caused by traditional manual detection methods, reduces costs and improves detection efficiency, meets the needs of existing large-scale hatcheries, and has great practical value. Attached Figure Description

[0021] Figure 1 This describes the working principle of the embryo screening method based on millimeter-wave radar in the embodiments of the present invention;

[0022] Figure 2 This is a flowchart illustrating the preprocessing of radar data described in various embodiments of the present invention;

[0023] Figure 3 This is a flowchart of the least squares circle fitting method for correlation optimization in an embodiment of the present invention. Detailed Implementation

[0024] The embodiments of the technical solution of the present invention will now be described in detail with reference to the accompanying drawings. These embodiments are only used to more clearly illustrate the technical solution of the present invention and are therefore merely examples, and should not be construed as limiting the scope of protection of the present invention.

[0025] To address the issues of existing embryo egg detection equipment requiring non-contact testing and the low accuracy and efficiency of embryo egg detection in hatcheries, this application provides an embryo egg screening method and device based on millimeter-wave radar. This method enables real-time and rapid detection of embryo egg development at any location in the early stages of incubation, allowing for timely screening of infertile eggs, dead embryos, and other embryo eggs, thereby improving incubation efficiency and reducing resource consumption.

[0026] This application discloses an embryo egg screening and detection device, including an embryo incubation device, a millimeter-wave radar motion detection device, a multi-point signal acquisition device, and a computer signal processing terminal. The embryo incubation device consists of an incubation box and an embryo egg fixing container. The millimeter-wave radar motion detection device consists of a millimeter-wave radar sensor, a robotic arm support, and a sliding guide rail. The robotic arm support is used to adjust the millimeter-wave radar sensor so that it is aligned with the center of the embryo incubation device. The sliding guide rail is used to move the millimeter-wave radar along the sliding guide rail, thereby realizing the detection and screening of all embryo eggs inside the embryo incubation device along the movement path. The multi-point signal acquisition device is used to receive data transmitted by the radar and send it to the computer signal processing terminal. The computer signal processing terminal is used to receive radar data and perform signal processing.

[0027] The working principle of the embryo screening method based on millimeter-wave radar provided by this invention is as follows: Figure 1 As shown:

[0028] The embryos to be screened are placed in the embryo incubation device. After the embryos are placed, the millimeter-wave radar sensor moves along the sliding guide rail. When the number of embryos entering the detection area reaches the threshold, the millimeter-wave radar sensor stops moving and detects each embryo in the detection area. After the detection is completed, the millimeter-wave radar sensor continues to move and detects the embryos in the next detection area.

[0029] Furthermore, the millimeter-wave radar sensor is configured in a one-transmit, four-receive mode, transmitting frequency-modulated continuous wave (FMCW) with a predetermined frequency of 30 GHz to 300 GHz, irradiating the embryonic egg with 400 frames of electromagnetic waves and receiving the echo signal, and then mixing the echo signal with the transmitted signal to obtain an intermediate frequency signal.

[0030] The electromagnetic waves irradiate poultry eggs within this frequency range and can penetrate the shells of different types of eggs, thereby detecting the embryonic development status. The poultry eggs include, but are not limited to, chicken eggs, duck eggs, goose eggs, and bird eggs. The shell thickness varies among different types of poultry, and the millimeter-wave radar motion detection device is designed to fully consider these differences. To accommodate these differences, the millimeter-wave radar system is optimized within a frequency range of 30 GHz to 100 GHz, selecting appropriate frequency bands to ensure that the electromagnetic waves can effectively penetrate various eggshells and provide sufficient resolution to detect embryonic development. The millimeter-wave radar sensor uses a high-frequency band to ensure penetration of thicker eggshells (such as duck and goose eggs) and to provide higher spatial resolution; while for thinner eggshells (such as bird eggs), a lower frequency band is sufficient for penetration. For chicken eggs, duck eggs, goose eggs, and bird eggs, frequency ranges of 60 GHz to 100 GHz, 50 GHz to 90 GHz, 40 GHz to 80 GHz, and 30 GHz to 60 GHz are selected respectively to optimize penetration and signal detection effects, ensuring accurate monitoring of embryonic development. The system is also set to a frame rate of 400 frames per second to ensure sufficient data collection during the detection process, thereby improving the accuracy of embryonic development detection.

[0031] The frequency-modulated continuous pulse emitted by a millimeter-wave radar sensor can be represented as:

[0032]

[0033] in, It is the amplitude of the transmitted signal. The slope of the sawtooth wave. The center frequency of the radar transmitted signal. This represents the initial phase of the transmitted signal.

[0034] The expression for the echo signal received by the millimeter-wave radar sensor is:

[0035]

[0036] in It is the amplitude of the echo signal. This is the delay of the radar echo signal;

[0037] The received echo signal and the transmitted signal are mixed, and then low-pass filtered to obtain the intermediate frequency (IF) signal. The expression for the IF signal is:

[0038]

[0039] Its frequency = phase = .

[0040] Furthermore, the intermediate frequency signal is transmitted through a signal acquisition device to a computer signal processing terminal for processing. The specific processing procedure is as follows:

[0041] 1. Perform preprocessing operations on the radar data to remove DC offset and suppress noise interference from clutter on the target echo;

[0042] 2. Perform range-FFT and angle-FFT on the preprocessed signal, use CFAR constant false alarm rate to perform 2D target recognition, and determine the embryonic egg target in the area;

[0043] 3. Extract the phase information of the embryonic egg target within the region and calculate the phase difference between adjacent distance compartments.

[0044] 4. Impulse noise was removed using a moving average filter, and the embryonic heartbeat component was extracted using an 8th-order IIR bandpass filter to obtain the number of embryonic heartbeats per minute;

[0045] Specifically, the following steps are included:

[0046] Step 1 includes the following: The radar data is subjected to a correlation-optimized least squares method for circular fitting to remove DC offset; the phasor mean cancellation algorithm is used to remove static clutter components from the radar data. The specific steps are as follows:

[0047] Complex data for each range gate is extracted from radar data. These complex data points are mapped onto a complex plane, with the real part as the abscissa and the imaginary part as the ordinate, resulting in a set of discrete complex plane points. To remove the DC component, a least-squares fitting method is applied to these points to establish a circle fitting model. The position of the circle's center is then determined using least-squares fitting. The radius of the circle. The coordinates of the circle's center represent the average DC offset of the complex signal. The formula for least squares circle fitting is as follows:

[0048]

[0049] in and Let be the coordinates of the center of the circle. , Let be the radius of the circle, and n be the total number of data points at the current distance gate. Given data points on the complex plane. The goal is to minimize the error function: minimizing the difference in distance from all data points to the fitted circle, so that the radius and center of the fitted circle best represent the data points. By minimizing this error function, the optimal estimates of the circle's center and radius can be obtained.

[0050] Based on least-squares circle fitting for each distance gate, the complex data of adjacent distance gates are further correlated using the Pearson correlation coefficient. The formula for calculating the Pearson correlation coefficient is as follows:

[0051]

[0052] Where r is the Pearson correlation coefficient. and For the complex data points of the adjacent distance gate, and The mean of the two values ​​is taken as the sample mean. If the Pearson correlation coefficient between adjacent distance gates is high, the fitting center of the two distance gates is replaced with the mean of the least square fitting centers of the two adjacent distance gates. This method can effectively reduce redundant calculation errors caused by high consistency between adjacent distance gates, and improve the smoothness and overall consistency of the fitting results.

[0053] The fitted center coordinates are used as the offset value. This offset is subtracted from the original complex data, thereby moving the center to the origin (0,0) on the complex plane. This effectively removes the DC component in the radar echo signal and introduces the correlation calculation of adjacent range gates, thereby improving the accuracy of target detection and parameter estimation and enhancing the radar system's ability to detect targets.

[0054] The distance from the stationary target to the radar antenna is constant, and the time delay of the stationary target on each received pulse is also constant. By averaging all received pulses, the reference received pulse can be obtained. Then, by subtracting the reference received pulse from each received pulse, the phasor mean cancellation result can be obtained, thus suppressing the interference of static clutter components.

[0055] Step 2 includes the following: performing range-FFT and angle-FFT on the sampling point data of each chirp of the preprocessed radar data to obtain the range-angle matrix, and making a preliminary estimate of the orientation of multiple embryonic egg targets.

[0056] A 2D-CFAR algorithm is used to achieve multi-target detection of embryonic eggs. A sliding window is defined to evaluate its local background noise characteristics, including: detection units, guard units, and reference units. Statistical analysis is performed on the signal intensity of the reference units to estimate the background noise characteristics of the area where the current detection unit is located. Combined with a preset false alarm probability threshold, a dynamic detection threshold is calculated using a formula: ,in The scaling factor is determined based on the false alarm probability (PFA). The background noise statistics of the reference unit are used; if the signal strength of the detection unit is greater than the threshold T, the detection unit is determined to be the target point.

[0057] Step 3 includes the following: Phase information of the embryonic egg target within the region is extracted using the phase arctangent. Due to the arctangent operation, the phase is generally limited to between -π and π. When some values ​​in the actual phase exceed this range, phase entanglement occurs. Therefore, phase untangling calculation is required for the extracted phase. Whenever the phase difference between consecutive values ​​is greater than or less than ±π, the untangled phase is obtained by subtracting 2π from the phase.

[0058] A phase difference operation is performed on the unwound phase by subtracting consecutive phase values. The specific formula is as follows: ,in The phase value at the current point. The phase value of the previous point. The phase difference value is achieved by continuously unwinding the phase of the current sampling point and subtracting it from the previous sampling point. This will help enhance the heartbeat signal and eliminate phase drift in the hardware receiver, eliminate interference from noise on the embryo's vital signs, and obtain a signal containing the heartbeat.

[0059] Step 4 includes the following: using a moving average filter with a window size of 5 to filter out high-frequency noise in the phase difference signal.

[0060] An IIR bandpass filter is applied to the phase difference signal after moving average filtering to filter out other signals and extract the heartbeat-related frequency components from the phase signal. Because the incubation process varies among different types of eggs, the time at which the embryonic heartbeat frequency and intensity reach their peak also differs at different days. Therefore, the stopband and passband cutoff frequencies of the bandpass filter need to be adjusted according to the specific incubation characteristics of different egg species to accurately extract the embryonic heartbeat signal. For example, the time when the heartbeat frequency and intensity of chicken, duck, and goose eggs reach their peak differs; adaptively selecting different filter frequency ranges helps to more accurately monitor and analyze embryonic development.

[0061] An FFT is performed on the bandpass-filtered signal, traversing the positive frequency portion of the spectrum to find the frequency component with the largest amplitude. A threshold is set; if the amplitude is greater than the threshold, the current frequency component is considered to have a valid heartbeat signal, and the embryo's heartbeats per minute (BPM1) are calculated. A time-domain peak detection algorithm is used to detect the peaks of the embryo's heartbeat signal and calculate its heartbeats per minute (BPM2). BPM1 and BPM2 are then averaged to obtain the accurate embryo's heartbeats per minute.

[0062] The viability of the tested embryos is determined based on the precise number of heartbeats per minute, the hatching time of the tested embryos, and the type of the tested embryos, for subsequent screening.

[0063] Currently, poultry egg embryo detection technology generally relies on manual inspection, which is highly inefficient and lacks specific quantitative standards. Furthermore, the accuracy of selection decreases with increasing working time. Currently, the detection methods used in domestic research on poultry egg embryo culture technology are mainly invasive measurement methods, which have strict requirements on the testing environment. To address these problems, this invention proposes a new solution, comprising an embryo incubation device, a millimeter-wave radar motion detection device, a multi-point signal acquisition device, and a computer signal processing terminal. This invention enables non-contact detection of various poultry egg embryo incubation states, effectively improving the accuracy and efficiency of detection, solving the problems of interference and potential damage to embryos caused by traditional manual inspection methods, reducing costs, meeting the needs of existing large-scale hatcheries, and possessing significant practical value.

[0064] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An embryo screening method based on millimeter-wave radar, characterized in that, Includes the following steps: S1: Control the millimeter-wave radar sensor to send a frequency-modulated continuous wave (FMCW) into the embryonic egg, receive the echo signal and mix it with the transmitted signal to obtain an intermediate frequency signal; S2: The mixed intermediate frequency signal data is sent to the signal processing terminal for signal processing through the signal acquisition device; S3: Preprocess the collected radar data; S4: Extract target phase information from the preprocessed signal in S3; S5: For the heartbeat component after bandpass filtering in S4, a time-domain peak-finding algorithm is used to find the peak of the embryonic heartbeat signal, and the frequency of the peak is calculated to obtain the embryonic heartbeat frequency; S6: Based on the embryonic heart rate described in S5, combined with the hatching time of the tested embryo and the type of embryo being tested, determine the survival status of the tested embryo. In step S3, the preprocessing of the radar data includes the following steps: S31: Perform range-FFT and angle-FFT on each chirp of the radar signal to convert the time-domain data into frequency-domain data; S32: The least squares method is used to perform circular fitting on the radar data to eliminate DC offset; S33: Employs a phasor mean cancellation algorithm to suppress static clutter components in radar data; S34: Construct a distance-angle map to perform two-dimensional localization of multiple embryonic eggs in the detection area, and use the constant false alarm rate (CFAR) algorithm to determine the number and location of embryonic eggs in the detection area; The step of performing circle fitting on radar data using the least squares method includes: S321: Extract the complex data of each range gate from the radar data, map the complex data points of each range gate onto the complex plane, and use the real part as the abscissa and the imaginary part as the ordinate to obtain a set of discrete complex plane points. S323: Perform least squares fitting on the complex data points of each distance gate to establish a circle fitting model; solve for the position of the circle center through least squares fitting. And the radius of the circle; S323: Based on the least squares circle fitting for each distance gate, the correlation of the complex data of adjacent distance gates is further calculated using the Pearson correlation coefficient; S324: If the Pearson correlation coefficient between adjacent distance gates is high, replace the fitted center of the two distance gates with the mean of the least squares fitted centers of the two adjacent distance gates. S325: The fitted center coordinates are used as the offset value. The offset is subtracted from the original complex data to remove the DC component in the radar echo signal. The formula for least squares circle fitting is as follows: in and Let be the coordinates of the center of the circle. , Let n be the radius of the circle, and n be the total number of data points at the current distance gate; given the data points on the complex plane. The goal is to minimize the error function: minimize the difference in distance from all data points to the fitted circle, so that the radius and center of the fitted circle best represent the data points; by minimizing this error function, the optimal estimates of the center and radius can be obtained; The formula for calculating the Pearson correlation coefficient is as follows: Where r is the Pearson correlation coefficient. and For the complex data points of the adjacent distance gate, and The sample mean of both. If the Pearson correlation coefficient between adjacent distance gates is high, then the fitting center of the two distance gates is replaced with the two adjacent distance gates. The mean of the least squares fitting circle center away from the gate; The embryo screening method based on millimeter-wave radar is implemented through a millimeter-wave radar-based embryo screening device. This device includes an embryo incubation device, a millimeter-wave radar motion detection device, a multi-point signal acquisition device, and a signal processing terminal. The millimeter-wave radar motion detection device consists of a millimeter-wave radar sensor, a robotic arm support, and a sliding guide rail. The robotic arm support is used to adjust the millimeter-wave radar sensor so that it faces the center of the embryo incubation device. The sliding guide rail is used to move the millimeter-wave radar along the sliding guide rail, thereby enabling the detection and screening of all embryos inside the embryo incubation device along the moving path.

2. The embryo screening method based on millimeter-wave radar according to claim 1, characterized in that, The embryo screening device selects the corresponding electromagnetic wave frequency range based on the eggshell thickness of different types of poultry eggs and the characteristics of the incubation process to optimize penetration ability and signal quality.

3. The embryo screening method based on millimeter-wave radar according to claim 2, characterized in that, For chicken eggs, duck eggs, goose eggs, and bird eggs, the selected frequency ranges are 60 GHz to 100 GHz, 50 GHz to 90 GHz, 40 GHz to 80 GHz, and 30 GHz to 60 GHz, respectively.

Images