An egg quality identification method, device and computer based on optical spectrum

Abstract

The present application relates to a method, device and computer for identifying the quality of eggs based on optical spectra. The method includes: identifying the physical characteristics of the eggshell of a target egg from a multi-dimensional captured image of the target egg to obtain the physical characteristic data of the eggshell of the target egg; planning the illumination of the target egg at a spatial angle of the light source according to the physical characteristic data of the eggshell to obtain the multi-dimensional light source illumination parameters of the target egg; controlling each modulation excitation source to perform time-resolved spectroscopy detection on the target egg according to the multi-dimensional light source illumination parameters to obtain the spectral detection data of each egg; and evaluating the health characteristics of the target egg according to the spectral detection data of each egg to obtain the egg quality data of the target egg. Using this method can effectively improve the accuracy and efficiency of egg quality identification, and further improve the degree of automation and industrial application value in industrial production.

Description

Technical Field

[0001] The present application relates to the field of intelligent detection technology, and particularly to an egg quality identification method, device and computer based on optical spectrum. Background Art

[0002] In traditional technology, a high-definition camera is used to collect the appearance image of an egg, and color, shape, crack and other characteristic features are extracted through image preprocessing (such as denoising, segmentation, feature extraction). Then, these features are input into a trained classification model (such as SVM, random forest or convolutional neural network CNN) to realize automatic discrimination of egg quality, such as identifying whether it is damaged, whether the surface is clean or grading the size. However, the traditional technology does not identify the internal situation of the egg, resulting in low accuracy of egg quality identification. Summary of the Invention

[0003] Based on this, in view of the above technical problems, it is necessary to provide an egg quality identification method, device and computer based on optical spectrum that can effectively improve the accuracy of egg quality identification.

[0004] In a first aspect, the present application provides an egg quality identification method based on optical spectrum, including:

[0005] Performing shell physical feature identification on a multi-dimensional captured image of a target egg to obtain shell physical feature data of the target egg;

[0006] According to the shell physical feature data, performing light source spatial angle irradiation planning on the target egg to obtain multi-dimensional light source irradiation parameters of the target egg;

[0007] According to the multi-dimensional light source irradiation parameters, controlling each modulation excitation source to perform time-resolved spectroscopy detection on the target egg to obtain each egg spectroscopy detection data;

[0008] According to each of the egg spectroscopy detection data, evaluating the egg health characteristics of the target egg to obtain egg quality data of the target egg.

[0009] In a second aspect, the present application further provides an egg quality identification device based on optical spectrum, including:

[0010] A feature identification module, configured to perform shell physical feature identification on a multi-dimensional captured image of a target egg to obtain shell physical feature data of the target egg;

[0011] An irradiation planning module, configured to perform light source spatial angle irradiation planning on the target egg according to the shell physical feature data to obtain multi-dimensional light source irradiation parameters of the target egg;

[0012] A spectral detection module, configured to control each modulation excitation source to perform time-resolved spectral detection on the target eggs according to the multi-dimensional light source irradiation parameters, and obtain spectral detection data of each egg;

[0013] A quality analysis module, configured to evaluate the egg health characteristics of the target eggs according to the spectral detection data of each egg, and obtain egg quality data of the target eggs.

[0014] In a third aspect, the present application further provides a computer, including a memory and a processor, where the memory stores a computer program, and when the processor executes the computer program, any step of an egg quality identification based on optical spectrum is implemented.

[0015] The above-mentioned method, device and computer for identifying egg quality based on optical spectrum collect and analyze multi-dimensional images of target eggs, accurately identify physical characteristic information of their eggshells, including surface texture, shape structure, color distribution, etc., so as to provide a basis for subsequent irradiation planning of the spatial angle of the light source, ensuring comprehensive and targeted light coverage; based on the obtained multi-dimensional light source irradiation parameters, the modulation excitation source can perform efficient time-resolved spectral detection on the target eggs, obtaining rich and high-precision spectral response data; further through in-depth analysis of these spectral detection data, internal health characteristics of the eggs such as freshness, protein quality, presence of cracks or foreign objects, etc. can be comprehensively evaluated, thereby outputting accurate and reliable egg quality data. It can effectively improve the accuracy and efficiency of egg quality identification, and further improve the automation degree and industrial application value in industrial production. Description of the Drawings

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required for use in the description of the embodiments or related technologies. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0017] Figure 1 It is an application environment diagram of a method for identifying egg quality based on optical spectrum in an embodiment;

[0018] Figure 2 It is a flowchart of a method for identifying egg quality based on optical spectrum in an embodiment;

[0019] Figure 3 It is a flowchart of a method for obtaining the first type of egg quality data in an embodiment;

[0020] Figure 4Schematic flow chart of a method for obtaining egg health characteristic data in an embodiment;

[0021] Figure 5 Schematic flow chart of a second method for obtaining egg quality data in an embodiment;

[0022] Figure 6 Schematic flow chart of a method for obtaining multi-dimensional light source irradiation parameters in an embodiment;

[0023] Figure 7 Schematic flow chart of a first method for obtaining the irradiation area of each light source in an embodiment;

[0024] Figure 8 Schematic flow chart of a second method for obtaining the irradiation area of each light source in an embodiment;

[0025] Figure 9 Schematic block diagram of an egg quality identification device based on optical spectrum in an embodiment;

[0026] Figure 10 Internal structure diagram of a computer device in an embodiment. Detailed implementation manners

[0027] In order to make the objectives, technical solutions and advantages of the present application clearer and more understandable, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0028] A method for identifying egg quality based on optical spectrum provided by an embodiment of the present application can be applied to an application environment as shown in Figure 1 . Among them, the terminal 102 communicates with the server 104 through a network. The data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104, or can be placed in the cloud or other network servers. Among them, the portable wearable device can be a smart watch, a smart bracelet, a head-mounted device, etc. The server 104 can be implemented by an independent server or a server cluster composed of multiple servers.

[0029] In an exemplary embodiment, as shown in Figure 2 , a method for identifying egg quality based on optical spectrum is provided. Taking the application of this method to the server in Figure 1 as an example, the method includes the following steps 202 to step 208. Among them:

[0030] Step 202, perform shell physical feature recognition on the multi-dimensional captured image of the target egg to obtain the shell physical feature data of the target egg.

[0031] Among them, the multi-dimensional captured image can be a set of image data obtained by comprehensively collecting the target eggs using imaging techniques with multiple angles, multi-spectra, and even multiple time points. It not only includes traditional visible light images but may also cover dimensional information such as infrared images, ultraviolet images, depth images, etc., for obtaining detailed features of the egg surface and local structures from multiple spatial and spectral dimensions.

[0032] Among them, the identification of eggshell physical characteristics can be the processing and analysis of multi-dimensional images to extract the characteristic information of the eggshell surface at the physical level, such as color distribution, texture structure, shape contour, stain distribution, crack conditions, spots, and pores.

[0033] Among them, the eggshell physical characteristic data can be the quantitative description results of the eggshell physical properties extracted from image analysis, usually represented in a structured form, including parameters such as texture direction, color mean, crack length, eggshell curvature, stain distribution, etc.

[0034] Specifically, a multi-dimensional imaging system (such as multi-angle visible light imaging, near-infrared imaging, etc.) is used to comprehensively capture the target eggs to obtain their multi-view image data; then image processing and analysis algorithms (such as edge detection, texture analysis, color distribution recognition, etc.) are used to extract the physical characteristic information of the eggshell, including the surface roughness, color uniformity, spot distribution, crack conditions, geometric morphology, stain distribution, etc. of the eggshell, thereby generating eggshell physical characteristic data.

[0035] Step 204: According to the eggshell physical characteristic data, perform a light source spatial angle irradiation plan for the target egg to obtain the multi-dimensional light source irradiation parameters of the target egg.

[0036] Among them, the light source spatial angle irradiation plan can be to formulate a light source irradiation scheme most suitable for the current egg structure based on the eggshell physical characteristic data, combined with three-dimensional modeling and optical simulation analysis.

[0037] Among them, the multi-dimensional light source irradiation parameters can be the output results of the irradiation planning process, including comprehensive control parameters of multiple light sources in terms of spatial coordinates, irradiation angles, luminous intensities, wavelength ranges, modulation frequencies, etc.

[0038] Specifically, a three-dimensional model is built based on the physical characteristic data of the eggshell to reconstruct the spatial geometric model of the target egg. Combining the texture distribution, color characteristics, and possible defect areas on the eggshell surface, the spectral response sensitivity of each area is analyzed to obtain a reasonable area for egg detection irradiation. On this basis, using the light propagation simulation algorithm, the light effects under different incident angles, light source positions, and light intensities are simulated and calculated to evaluate the influence of different irradiation parameters on the spectral detection quality of the eggs. According to the simulation results, a set of optimal light source configuration schemes is selected to determine parameters such as the spatial position, irradiation direction, irradiation angle, and light intensity distribution of each modulation excitation source, thus forming multi-dimensional light source irradiation parameters.

[0039] Step 206: According to the multi-dimensional light source irradiation parameters, control each modulation excitation source to perform time-resolved spectral detection on the target egg to obtain each egg spectral detection data.

[0040] Among them, the modulation excitation source can be a light source device with controllable light-emitting ability, such as an adjustable wavelength LED, a laser, or other modulation-type light sources, which can emit light of a specific wavelength at different times, frequencies, or intensities according to external instructions, and is used to excite the internal or surface materials of the egg to generate a spectral response.

[0041] Among them, time-resolved spectral detection can be a spectral acquisition technology, which not only acquires spectral data at different wavelengths but also records the dynamic change process of the spectrum at different time points.

[0042] Among them, the egg spectral detection data can be the spectral response results obtained by the time-resolved spectral detection system under specific irradiation conditions, usually including multiple spectral dimension information such as reflectivity, transmittance, absorbance, or fluorescence intensity, which reflects the physical and chemical states of the egg under specific wavelength excitation.

[0043] Specifically, according to the multi-dimensional light source irradiation parameters, control each modulation excitation source (such as LED light sources, lasers, etc. with different wavelengths) to start sequentially according to the specified spatial position and irradiation angle, and perform synchronous excitation according to the set time sequence to ensure multi-period and hierarchical light excitation of the target egg. During the excitation, the time-resolved spectral detection element starts to collect the spectral response signals of the egg synchronously in real time, capturing the spectral changes such as reflection, transmission, and fluorescence of the egg material at different time points, thereby constructing each egg spectral detection data covering the time dimension, spectral dimension, and spatial response dimension.

[0044] Step 208: According to each egg spectral detection data, evaluate the egg health characteristics of the target egg to obtain the egg quality data of the target egg.

[0045] Among them, the healthy characteristics of eggs can be the index status of eggs themselves at the physiological or quality level, including freshness, internal structure integrity, presence of foreign objects or cracks, ratio and uniformity of egg white and egg yolk, moisture content, etc.

[0046] Among them, the egg quality data can be the comprehensive quantitative evaluation results obtained based on the evaluation of the healthy characteristics of eggs, usually presented in the forms of quality data, scoring data, whether it is qualified, processing suggestion data, etc., and is used to guide the grading, packaging, transportation or elimination processing of eggs, and support the industrialized egg quality management process.

[0047] Specifically, preprocess the spectral detection data of each egg, including operations such as noise filtering, baseline correction, spectral normalization, etc., and then use feature extraction algorithms to extract key spectral feature parameters from the preprocessed spectral detection data of each egg, such as reflectivity in specific bands, absorption peak positions, fluorescence intensity changes, etc.; based on the existing healthy characteristic evaluation models (such as support vector machines, neural networks or discriminant analysis, etc.), match and discriminate the extracted spectral features with healthy indicators of eggs such as freshness, protein content, egg yolk status, presence of cracks or impurities, etc.; on this basis, combined with the set quality evaluation standard system, conduct quantitative analysis on eggs to generate egg quality data, including egg health scores, egg quality data and corresponding recommended processing data, etc.

[0048] In the above method for identifying egg quality based on optical spectra, by collecting and analyzing multi-dimensional images of the target egg, accurately identify the physical characteristic information of its eggshell, including surface texture, shape structure, color distribution, etc., so as to provide a basis for the subsequent illumination planning of the light source spatial angle, ensuring that the illumination coverage is comprehensive and targeted; based on the obtained multi-dimensional light source illumination parameters, the modulation excitation source can perform efficient time-resolved spectral detection on the target egg to obtain rich and high-precision spectral response data; further through in-depth analysis of these spectral detection data, the internal healthy characteristics of eggs such as freshness, protein quality, presence of cracks or foreign objects, etc. can be comprehensively evaluated, so as to output accurate and reliable egg quality data. It can effectively improve the accuracy and efficiency of egg quality identification, and further improve the automation degree and industrial application value in industrial production.

[0049] In an exemplary embodiment, as Figure 3 shown, according to the spectral detection data of each egg, evaluate the healthy characteristics of the target egg to obtain the egg quality data of the target egg, including steps 302 to 306. Among them:

[0050] Step 302, perform time decay analysis on the spectral detection data of each egg respectively to obtain each decay analysis feature vector.

[0051] Among them, time decay analysis can be a technology for modeling and analyzing the process of the response signal of eggs changing with time during spectral detection. By observing how the intensity of the spectral signal at different wavelengths decays with time, it can reflect the changes in the absorption, reflection, or scattering characteristics of the internal tissues of eggs to light stimulation, thereby revealing the activity, stability, and structural integrity of egg materials.

[0052] Among them, the decay analysis eigenvector can be a set of vectors composed of key parameters extracted from time decay analysis, used to quantify the time response characteristics of eggs under spectral excitation. This vector may include multiple dimensions such as initial light intensity, half-life time, maximum decay rate, signal-to-noise ratio, etc.

[0053] Specifically, perform time series analysis on the obtained egg spectral detection data, and pay special attention to the decay behavior of the spectral intensity with time under different wavelength excitations. By establishing a time decay curve and using fitting algorithms (such as exponential decay model, polynomial regression, etc.) to model each group of spectral data, extract key parameters describing the change speed, amplitude, and stability of the spectral response, such as initial intensity, half-life time, decay constant, etc., and then construct the decay analysis eigenvector corresponding to each detection channel. The decay analysis eigenvector can be used to characterize the dynamic response characteristics of egg materials to spectral stimulation.

[0054] Step 304, analyze the health characteristics of the target egg according to each decay analysis eigenvector to obtain egg health characteristic data.

[0055] Among them, the health characteristics can be internal indicators reflecting the overall quality and edible safety of eggs, including freshness, physical structure states of egg white and yolk, cracks or impurities, internal moisture content, etc.

[0056] Among them, the egg health characteristic data can be a set of quantitative information that can specifically describe the health status of eggs generated based on the evaluation of the decay analysis eigenvector, including indicators such as freshness data, structural integrity index, internal anomaly probability, etc., usually represented in the form of structured data.

[0057] Specifically, input each decay analysis eigenvector into a pre-trained health characteristic evaluation model. The health characteristic evaluation model establishes a mapping relationship between decay characteristics and the internal state of eggs based on machine learning methods (such as support vector machines, random forests, or deep neural networks), performs classification or regression analysis on each input group of decay analysis eigenvectors, and identifies physiological indicators such as freshness level, structural integrity, and states of egg white and yolk corresponding to them in historical samples; extract multiple quantitative indicators related to health according to the analysis results, such as spectral stability data, crack potential data, freshness data, etc., and structurally integrate these indicators into a dataset reflecting the current internal health status of eggs, that is, egg health characteristic data.

[0058] Step 306: Predict the healthy development trend of the egg health characteristic data to obtain the egg quality data.

[0059] Among them, the healthy development trend can be the change trajectory and development direction that the egg health characteristic data may present in a future period of time. This trend is obtained through predictive analysis of the current health characteristic data and the historical change model, and is used to judge the quality change speed of the egg during storage or transportation, whether it will accelerate spoilage or remain stable.

[0060] Specifically, organize the egg health characteristic data according to the time series structure, and combine the known egg storage rules and environmental parameters (such as temperature, humidity, light, etc.) to construct the input features of the prediction model. Use time series prediction algorithms, such as ARIMA model, LSTM neural network or ensemble learning method, to model and predict the change trends of various health indicators (such as fresh data, structural integrity data, etc.) in a future period of time, and evaluate the possible quality decline, spoilage risk or structural degradation over time; splice the predicted health change trend data with the current characteristic state to output the egg quality data.

[0061] In this embodiment, by performing time decay analysis on the egg spectral detection data and extracting the dynamic response feature vectors, it can more accurately reflect the spectral change law of the egg at different time points and improve the perception ability of its internal state change; then, based on these decay feature vectors, perform health characteristic analysis to achieve quantitative evaluation of key indicators such as egg freshness, structural integrity and biological activity; further, by predicting the development trend of the health characteristic data, it is possible to predict in advance the quality evolution of the egg during storage and transportation, so as to output scientific and reliable egg quality data. It not only improves the intelligence and accuracy of egg quality evaluation, but also has an early warning function, providing significant technical support for grading processing, shelf life management and supply chain optimization, and has good practical application value and industrial promotion prospects.

[0062] In an exemplary embodiment, as Figure 4 shown, analyze the health characteristics of the target egg according to each decay analysis feature vector to obtain the egg health characteristic data, including steps 402 to 408. Among them:

[0063] Step 402: Integrate each decay analysis feature vector according to the time identifier in each decay analysis feature vector to obtain a decay data integration matrix.

[0064] Among them, the attenuation data integration matrix can be a multi-dimensional data structure formed by integrating multiple attenuation analysis eigenvectors in chronological order and feature dimensions, which is used to comprehensively represent the spectral response changes of eggs at different time points and different bands during spectral detection.

[0065] Specifically, time identification information is extracted from each attenuation analysis eigenvector, and all vectors are normalized and aligned in chronological order. Then, the data with the same time label are stacked by dimension to construct a multi-dimensional feature matrix reflecting the evolution of the attenuation process over time - that is, the attenuation data integration matrix. This attenuation data integration matrix still comprehensively retains the response characteristics of the target egg under different spectral conditions and different time points.

[0066] Step 404: Perform multi-dimensional local feature extraction on the attenuation data integration matrix according to the health impact characteristic factors of the target egg to obtain local attenuation characteristic data.

[0067] Among them, the health impact characteristic factors can be key spectra or time attenuation characteristics that are representative and judgmental for the health status of eggs, such as attenuation rate, instantaneous amplitude fluctuation, energy concentration region, peak shift, etc.

[0068] Among them, multi-dimensional local feature extraction can be a process of extracting fine-grained features with local representativeness for a specific time window and spectral dimension region in the attenuation data integration matrix.

[0069] Among them, the local attenuation characteristic data can be a structured feature set extracted from the attenuation matrix, which reflects the dynamic response characteristics within a specific time period or spectral interval. It describes information such as signal intensity change, fluctuation frequency, and attenuation form in the local area, and is an important basis for measuring whether there are abnormalities in the internal micro-state of eggs.

[0070] Specifically, based on the known health impact characteristic factors closely related to egg health, such as spectral attenuation rate, local amplitude change, instantaneous response difference, etc., determine the time period and spectral dimension region that need to be analyzed key points; then, by setting a sliding window or using a multi-scale analysis method, scan and segment a specific local area in the attenuation data integration matrix, and extract the fine-grained attenuation characteristics shown by each area in the time and frequency dimensions; apply local feature extraction algorithms (such as wavelet transform, short-time Fourier transform, or convolution operation) to model each selected area, and refine key indicators that can reflect changes in health status, such as local energy distribution, mutation point position, or signal instability, etc.; these local attenuation characteristic data extracted from different areas.

[0071] Step 406: Perform dimensionality reduction and fusion on the local attenuation characteristic data to obtain an attenuation characteristic supervector.

[0072] Among them, dimensionality reduction and fusion can be a process of compressing multiple high-dimensional local attenuation feature data into a low-dimensional space through a dimensionality reduction algorithm and further performing information fusion.

[0073] Among them, the attenuation feature supervector can be a unified and highly expressive low-dimensional feature vector formed after dimensionality reduction and fusion. It synthesizes the key information of multiple local features and can comprehensively and accurately characterize the spectral attenuation characteristics of eggs during the entire detection period.

[0074] Specifically, perform normalization processing on each local attenuation feature data, and apply dimensionality reduction techniques such as principal component analysis (PCA), linear discriminant analysis (LDA), or deep learning-based autoencoders to compress the features of each normalized local attenuation feature data, remove redundant information, and retain the principal component features that contribute most to health discrimination; through methods such as feature splicing, weighted averaging, or feature attention mechanisms, fuse the dimensionality-reduced local attenuation feature data to integrate and form a unified and highly expressive low-dimensional vector, obtaining an attenuation feature supervector representing the overall health status of the eggs.

[0075] Step 408, input the attenuation feature supervector into the health characteristic analysis model of the target egg to obtain egg health feature data.

[0076] Among them, the health characteristic analysis model can be an intelligent evaluation model for analyzing the mapping relationship between the attenuation feature supervector and the health status of eggs, usually constructed based on machine learning or deep learning methods.

[0077] Specifically, transfer the attenuation feature supervector as input data to a pre-constructed and trained egg health characteristic analysis model. This egg health characteristic analysis model can be designed based on machine learning algorithms (such as support vector machines, random forests) or deep learning architectures (such as multi-layer perceptrons, convolutional neural networks) to identify and analyze the complex mapping relationship between different feature combinations and health status; the egg health characteristic analysis model performs feature discrimination and state classification on the attenuation feature supervector, identifies the corresponding health data in the training samples, and analyzes situations such as structurally abnormal data, quality deterioration data, and freshness data; the egg health characteristic analysis model converts the analysis results into structured egg health feature data.

[0078] In this embodiment, by introducing time identifiers to integrate each attenuation analysis eigenvector, an attenuation data integration matrix with high time series accuracy is constructed, retaining the complete information of the egg spectral signal changing over time. Further, by combining health impact characteristic factors, multi-dimensional local feature extraction is performed on this matrix to effectively capture minute but crucial spatio-temporal signal features, enhancing the sensitivity and accuracy of health status recognition. Then, dimensionality reduction and fusion are performed on the local feature data to generate an attenuation feature supervector with high information density and low redundancy, enabling the analysis model to have stronger generalization ability and operation efficiency. Finally, a health characteristic analysis model is used to intelligently analyze the supervector, outputting accurate and quantifiable egg health characteristic data. This improves the automation and intelligence level of the egg health assessment process, has higher recognition accuracy and prediction value, and is widely applicable to modern egg quality control and intelligent screening systems.

[0079] In an exemplary embodiment, as Figure 5 shown, predicting the healthy development trend of egg health characteristic data to obtain egg quality data includes steps 502 to 506. Among them:

[0080] Step 502, identifying the physical analysis data and biological analysis data of the target egg from the egg health characteristic data.

[0081] Among them, the physical analysis data can be quantified information extracted from the external structure or surface characteristics of the egg, usually from physical perception means such as image processing, structured light measurement, or spectral reflection characteristics.

[0082] Among them, the biological analysis data can be quantified indicators obtained through time-resolved spectroscopy, fluorescence response, etc., for reflecting the internal physiological and chemical states of the egg. These data cover aspects such as freshness, protein degradation, moisture content, internal air chamber changes, and microbial activity, and can dynamically reveal potential quality changes of the egg during storage or transportation.

[0083] Specifically, through the feature labels, data source attributes, and historical sample correspondence relationships of the egg health characteristic data, various features are classified and identified. According to the physical sources and perception methods of the classified and identified features, data reflecting the external structure state of the egg (such as eggshell integrity, surface roughness, crack index, spectral reflection characteristics, etc.) are extracted as physical analysis data, which are usually static structure indicators from image analysis or spectral signals. At the same time, indicators representing changes in the internal components, physiological activities, and micro-metabolism of the egg (such as freshness score, moisture content trend, protein decay signal, volatile substance response, etc.) are identified from the remaining features and classified as biological analysis data. Most of these features come from dynamic spectral analysis and time response characteristics.

[0084] Step 504, predicting the biological abnormal development of the target egg based on the physical analysis data and the biological analysis data, and obtaining the biological evolution prediction data of the target egg.

[0085] Among them, biological abnormal development can be the process in which the internal biological characteristics of eggs (such as freshness, protein status, moisture, etc.) undergo abnormal changes during storage, transportation or environmental changes, which is manifested as indicators deviating from the normal evolution trajectory, such as a rapid decrease in freshness, a sharp increase in volatiles, or abnormal fluctuations in microbial indicators.

[0086] Among them, the biological evolution prediction data can be the time series data generated by deducing the changing trend of the biological characteristics of eggs in the future through a prediction model. It reflects the evolution path of key biological indicators (such as moisture, protein status, air chamber volume, etc.) of eggs under set storage conditions over time.

[0087] Specifically, first identify the key features in the biological analysis data that are highly correlated with quality changes, such as protein degradation rate, water evaporation trend, air chamber expansion rate and dynamic changes of volatile components; then input these time-related biological features and physical analysis data into the constructed biological abnormality development prediction model. The biological abnormality development prediction model is built based on time series algorithms (such as LSTM, GRU) or multivariate regression models to simulate the biological change process of eggs under specific storage conditions (such as temperature, humidity) under the current physical analysis data and these time-related biological features (such as changes in egg quality caused by gas exchange inside and outside the eggs); the biological abnormality development prediction model predicts the changing trends of these key indicators of the target eggs in the future time period by learning the evolution trajectory of historical sample data, and identifies whether there are potential development signs such as a sharp drop in freshness data, internal deterioration or abnormal microbial activity; finally, the above prediction results are output in the form of a structured time series to form biological evolution prediction data of the target eggs.

[0088] Step 506, analyzing the health trend of the target eggs according to the biological evolution prediction data to obtain egg quality data.

[0089] Among them, the health trend can be the development direction and change range of the overall health status of eggs in the future, which is usually based on a comprehensive analysis of current health characteristic data and biological evolution prediction data.

[0090] Specifically, the biological evolution prediction data is compared and analyzed with the current health characteristic data to identify the changing trends of key indicators (such as freshness data, moisture retention rate, internal structure integrity, etc.) in the time dimension; and based on a preset health development assessment model or threshold rule, analyze the data of the target eggs in the future healthy stable period, quality decline period or abnormal evolution period, and evaluate the remaining shelf life, risk change data in the future period, etc.; then combine the storage environment parameters (such as temperature, humidity) involved in the prediction process for multi-scenario deduction, and correct the health trend evolution path of eggs under different logistics or storage conditions; finally, generate a set of egg quality data including the current quality grade, future edibility estimate, potential risk warnings and recommended handling suggestions, etc.

[0091] In this embodiment, by accurately identifying the physical analysis data and biological analysis data from the egg health characteristic data, a comprehensive separate modeling of the surface structure characteristics and internal physiological state of the eggs is achieved, which helps to more finely grasp the quality composition elements of the eggs; further combining these two types of data to carry out biological abnormal development prediction can effectively capture potential abnormal evolution trends such as freshness decline, protein decomposition, and structure variation, and early warning of quality risks; finally, through the dynamic trend analysis of the biological evolution prediction data, the egg quality data covering the current state and future change direction is output, making the quality assessment more forward-looking and dynamic. It improves the scientificity and operability of egg quality prediction, is applicable to various application scenarios such as intelligent grading, shelf life management and cold chain logistics monitoring, and has significant practical application value and industrial promotion prospects.

[0092] In an exemplary embodiment, as Figure 6 shown, according to the eggshell physical characteristic data, a light source spatial angle irradiation plan is carried out for the target eggs to obtain the multi-dimensional light source irradiation parameters of the target eggs, including steps 602 to 606. Among them:

[0093] Step 602, according to the eggshell physical characteristic data, analyze the light source irradiation range of the target eggs to obtain the light source irradiation areas of the target eggs.

[0094] Among them, the light source irradiation range analysis can be an analysis process of determining which areas can be effectively irradiated and produce a stable spectral response at different light source incident angles by simulating and calculating the three-dimensional structure, surface reflection characteristics and optical behavior of the egg according to the eggshell physical characteristic data.

[0095] Among them, the light source irradiation area can be the sub-area of the egg surface selected through the irradiation range analysis and suitable for spectral detection. These areas have good geometric visibility and spectral response characteristics, and can generate clear and stable detection signals under the irradiation of a light source with a set angle and intensity, and are the core action areas for spectral excitation and detection.

[0096] Specifically, three-dimensional reconstruction technology is used to construct a spatial model of the egg surface based on the physical characteristic data of the eggshell, and its physical properties such as shape contour, surface curvature, texture direction and color distribution are clarified; further, the light propagation simulation algorithm is used, combined with the optical characteristics of the egg surface such as reflectivity and absorptivity, to simulate and analyze the path and response of light irradiating the egg surface at different angles and directions; based on the simulation results, it is identified which areas are prone to shadow obstruction, reflection interference or information loss, and these areas are eliminated, and areas with good spectral response, smooth surface and stable characteristics are preferentially retained as the target of the light source; based on the screening results of the areas, the egg surface is divided into multiple independent light source irradiation areas.

[0097] Step 604, controlling the egg cleaning gun head corresponding to the target egg to clean each light source irradiation area to obtain each clear irradiation area of the target egg.

[0098] Among them, the egg cleaning gun head can be an automated equipment component used to accurately clean specific areas on the surface of eggs, and is usually integrated into the pretreatment module of the detection system.

[0099] The clear irradiation area may be an irradiation area whose surface is clean, stain-free, and light-free after being directional cleaned by the egg cleaning gun tip.

[0100] Specifically, due to the storage conditions of eggs, it is not possible to fully clean the entire egg, so the three-dimensional coordinates, size and surface state information of each irradiation area are transmitted to the cleaning control module, and the appropriate cleaning parameters, including nozzle type, cleaning angle, water pressure intensity and duration, are intelligently matched according to the size of the irradiation area, the type of surface attachments (such as dust, water stains, stains) and the degree of interference they may cause to spectral reflection. Then the control system starts the corresponding egg cleaning gun head, and performs fixed-point, directional and high-precision cleaning of each light source irradiation area in a predetermined order to ensure that the surface structure of other areas is not affected; during the cleaning process, the cleaning effect is detected in real time through visual or reflectivity sensors, and secondary compensation is performed on the areas that are not thoroughly cleaned. Finally, when all irradiated areas are clean, residue-free, water film controllable, and light penetrable, they are marked as clear irradiation areas, providing a clean and stable optical response surface for subsequent high-quality lighting angle and intensity planning.

[0101] Step 606, performing light source illumination angle analysis and light source illumination intensity analysis on each clear illumination area, and obtaining multi-dimensional light source illumination parameters of the target egg.

[0102] Among them, the light source irradiation angle analysis can be to perform light simulation and reflection simulation on each clear irradiation area, evaluate the effect of light under different incident angles, and determine the optimal light incident direction. This analysis ensures that the light stimulates the internal response of the material in the most favorable way, reducing signal loss or reflection interference.

[0103] Among them, the light source irradiation intensity analysis can be a process of evaluating the optimal irradiation power and modulation frequency of the required light source based on the spectral response sensitivity of the irradiation area, the material absorption characteristics, and the target detection depth. It ensures that under the set angle conditions, the light source can output an appropriate energy level to stimulate sufficient signal intensity, while avoiding interference or signal saturation caused by over-irradiation.

[0104] Specifically, sample the surface state of each clear irradiation area to obtain its current reflection characteristics, surface normal direction, and local curvature information. Based on the eggshell surface model and optical characteristics, use light simulation and reverse ray tracing algorithms to perform multi-angle simulation irradiation analysis on each clear irradiation area, and evaluate the reflectivity, scattering pattern, and signal response intensity of light under different incident angles; through comparative analysis, determine the optimal irradiation angle for each area, that is, the incident direction that can maximize the excitation of spectral signals and minimize reflection interference. Then, according to the spectral response sensitivity and target detection depth of each clear irradiation area, combined with historical irradiation experience and physical models, match and optimize the required light source intensity, including irradiation power, modulation frequency, and wavelength range, and summarize the optimal angle parameters and irradiation intensity parameters of each clear irradiation area to generate a complete multi-dimensional light source irradiation parameter set.

[0105] In this embodiment, by performing light source irradiation range analysis on the eggshell physical characteristic data, the key areas on the egg surface that are most suitable for spectral detection can be accurately identified, avoiding light interference caused by abnormal curvature or surface defects; further, by controlling the egg cleaning nozzle to perform fixed-point cleaning on each irradiation area, surface stains and reflection interference objects can be effectively removed, ensuring that the irradiation area has ideal cleanliness and optical response conditions; then, perform irradiation angle and intensity analysis on the cleaned area, and accurately calculate the optimal irradiation parameters based on the surface structure and material penetration characteristics, realizing personalized and high-precision light source configuration. It ensures the stability of spectral excitation and the accuracy of detection data, significantly improves the reliability of subsequent spectral acquisition and egg quality analysis, and provides strong technical support for realizing intelligent, non-contact, and high-sensitivity egg detection.

[0106] In an exemplary embodiment, as Figure 7 shown, according to the eggshell physical characteristic data, perform light source irradiation range analysis on the target egg to obtain the light source irradiation areas of the target egg, including steps 702 to 706. Among them:

[0107] Step 702, clearing the stain feature data in the eggshell physical feature data to obtain eggshell contour feature data.

[0108] Among them, the stain feature data can be feature information related to the attachments on the surface of the egg, including dust, water stains, oil spots, foreign matter, etc., identified in the eggshell image or spectral information.

[0109] Among them, the eggshell contour feature data can be the real eggshell geometric structure information extracted from the image or sensor information after stains are removed, mainly including the edge contour, surface curvature, normal direction, geometric continuity and local texture distribution of the eggshell.

[0110] Specifically, the surface information contained in the physical feature data of the eggshell is preprocessed to identify the feature data representing interference factors such as stains, dust, and water stains, which are usually manifested as areas of abnormal reflection, color mutation, or discontinuous texture; these stain features are located and removed through algorithms such as image filtering, threshold segmentation, and reflection suppression, and are removed from the overall feature data. To ensure data integrity, the edge missing areas that may appear after cleaning are interpolated or reconstructed to restore the continuity of the eggshell structure; ultimately, only valid features that reflect the true geometric shape and surface structure are retained, such as edge contour lines, surface curvature distribution, texture direction, etc., to construct clean and accurate eggshell contour feature data.

[0111] Step 704, calculating the irradiation curvature range of the target egg according to the illumination wavelength parameter and the eggshell color parameter of the target egg.

[0112] Among them, the illumination wavelength parameter can be the working wavelength or wavelength range of the light source used to irradiate the eggs. Light of different wavelengths has different reflection, absorption and penetration behaviors on the surface and inside of the eggs.

[0113] Among them, eggshell color parameters can be quantitative data that describe the color distribution on the egg surface, usually including RGB values, spectral reflectance, color uniformity and color classification results. Eggshells of different colors have different absorption and reflection capabilities for light of different wavelengths.

[0114] The irradiation curvature range may be a curvature value interval most suitable for light source irradiation calculated based on factors such as the eggshell surface curvature, color response characteristics, and light wavelength.

[0115] Specifically, after obtaining the eggshell contour feature data, call the illumination wavelength parameters corresponding to the current detection task to clarify the wavelength range of the light source used and its reflection and absorption characteristics on different material and color surfaces. Subsequently, based on the color parameters of the eggshell surface, combined with the spectral response curves of each color at specific wavelengths, analyze the reflectivity and penetration ability of each color region to the illumination wave. Integrate the color response characteristics with the curvature information of the eggshell contour, simulate the energy distribution and reflection efficiency when light is incident at different angles on regions with different curvature radii, and identify which curvature ranges can generate stable and usable spectral signals under the current light source conditions. Based on these simulation results, establish a set of numerical boundaries and output a set of illumination curvature ranges for screening those eggshell surface regions that can not only ensure good light incidence but also facilitate subsequent signal acquisition.

[0116] Step 706, select the illumination regions of each light source from the eggshell contour feature data according to the illumination curvature range.

[0117] Specifically, after completing the calculation of the illumination curvature range, use this range as a screening condition and apply it to the previously obtained eggshell contour feature data. By matching the local curvature values of each surface region, identify the candidate regions that meet the illumination curvature requirements. Further conduct a geometric visibility analysis on these candidate regions to evaluate whether there are problems such as occlusion, shadow, or excessive incident angle under the light source layout of the detection system, and eliminate the regions where light is inaccessible or the light effect is poor. And conduct an optical performance evaluation on the remaining regions, and determine whether they can obtain ideal reflection or transmission signals under the preset wavelength conditions according to the surface smoothness and the matching degree between the normal direction and the light source direction. Mark all the regions that pass the screening and evaluation as the formal light source illumination regions, and assign accurate spatial coordinates, dimension information, and candidate illumination angle ranges to each region.

[0118] In this embodiment, by clearing the stain feature data in the eggshell physical feature data, effectively eliminating the surface interference factors that affect the accuracy of optical detection, and improving the clarity and reliability of subsequent image and spectral analysis. After further extracting the eggshell contour feature data, combined with the illumination wavelength parameters and the eggshell color parameters, according to the response characteristics of different materials and colors to specific wavelength light, accurately calculate the optimal illumination curvature range. Finally, select the most suitable light source illumination regions for spectral excitation from the eggshell contour according to this curvature range, effectively ensuring the reflection stability of the incident light and the response efficiency of the detection region. It realizes the automatic and high-precision illumination planning from feature extraction, condition matching to region screening, significantly improves the accuracy and adaptability of egg spectral detection, and provides solid data and physical support for the intelligent detection system.

[0119] In an exemplary embodiment, as Figure 8As shown in the figure, the light source irradiation angle and the light source irradiation intensity are analyzed for each clear irradiation area respectively to obtain the multi-dimensional light source irradiation parameters of the target egg, including steps 802 to 808. Among them:

[0120] Step 802, for any clear irradiation area, select the point with the smallest eggshell roughness in the clear irradiation area as the target irradiation point.

[0121] Among them, the eggshell roughness can be a quantitative index of the microscopic unevenness of the eggshell surface, usually evaluated by the average deviation of the surface height change (such as the Ra value). The smaller the roughness, the smoother the surface, the more stable the reflection of light, and the less scattering, which helps to improve the accuracy of spectral detection.

[0122] Among them, the target irradiation point can be the optimal light action point selected by analyzing the position with the smallest surface roughness in each clear irradiation area. This point usually has the characteristics of a flat surface, small reflection interference, and stable signal response, and is the basic coordinate for spectral excitation and energy control calculation.

[0123] Specifically, call the high-resolution imaging data or laser point cloud data to obtain the surface texture and topography information of each tiny point in each area, and then use the roughness calculation algorithm (such as the Ra or RMS method based on height deviation) to evaluate the roughness of all points one by one to generate the roughness distribution map of each area. Further, locate the point with the smallest roughness value in the roughness distribution map of each area, that is, the flattest and smoothest position on the surface. This point has the best reflection stability and the smallest scattering interference. Confirm this point as the target irradiation point of this clear irradiation area, and record its key information such as spatial coordinates and surface normal direction.

[0124] Step 804, according to the light wavelength parameter and the penetration material information of the target egg, perform a reflection analysis on the light source irradiation angle to obtain the target irradiation angle.

[0125] Among them, the penetration material information can be a data set describing the material characteristics encountered by light during the process of penetrating the egg structure, including eggshell thickness, refractive index, absorption coefficient, scattering coefficient, medium composition, etc.

[0126] Among them, the reflection analysis can be a process of simulating and evaluating the reflection intensity, directionality, and effectiveness of light after irradiating the target point at different angles.

[0127] Among them, the target irradiation angle can be the optimal light incident direction obtained after the reflection analysis, usually expressed by the pitch angle and the azimuth angle. This angle can ensure that the light generates a minimized reflection response at the target irradiation point and meets the acquisition conditions of the system detector.

[0128] Specifically, for each region, based on the spatial coordinates and surface normal direction of the target irradiation point, combined with the preset light wavelength parameters and the penetration material information of the target egg (such as eggshell, egg white, egg yolk, interlayer structure, etc.) as input conditions, based on the physical optical reflection model (such as Fresnel reflection law or bidirectional reflectance distribution function BRDF), simulate the light reflection behavior at different incident angles, analyze the reflectivity, reflection direction stability and energy loss of light on the surfaces of various penetration material information, evaluate the reflection efficiency and signal response quality at various angles, select the angle that can generate the weakest effective reflection signal as the target irradiation angle, and record the three-dimensional incident direction vector of this angle as the angle parameter in the optimal light source configuration scheme for this irradiation point.

[0129] Step 806, perform a transmission analysis on the light source irradiation intensity according to the light wavelength parameters and the penetration material information of the target egg to obtain the target irradiation intensity.

[0130] Among them, the transmission analysis can be a process of modeling and calculating the energy attenuation and propagation depth of light during the penetration of media such as the eggshell based on the wavelength of light, irradiation angle, and penetration material information.

[0131] Among them, the target irradiation intensity can be the minimum or appropriate light source output power required to achieve the desired detection depth and signal intensity under specific irradiation angles and material conditions.

[0132] Specifically, based on the target irradiation angle and the spatial position of the target irradiation point as constraint conditions, according to the preset light wavelength parameters and the information of the egg penetration material (such as the absorption coefficient, scattering coefficient, refractive index, etc. of the eggshell material), call the optical transmission model (such as Lambert-Beer law) to simulate the energy attenuation process of light penetration in the penetration material information, calculate the transmission threshold that can generate an effective spectral response in the internal detection region under different light intensity conditions; further deduce the minimum or appropriate light energy required to ensure penetration to the required depth and excite a stable detection signal at the target irradiation angle, so as to determine the optimal light source output parameters, including power density, modulation frequency, wavelength bandwidth, etc. Output these parameters as the target irradiation intensity corresponding to this point.

[0133] Step 808, fuse each target irradiation angle and each target irradiation intensity to obtain multi-dimensional light source irradiation parameters.

[0134] Specifically, after separately determining the target irradiation angle and target irradiation intensity corresponding to each clear irradiation area, they are structured and organized as a set of data for the light source arrangement. Then, based on the spatial distribution and lighting requirements of each point, the light source arrangement strategy is coordinated and optimized to ensure that the angle and intensity combinations in different areas are achievable, conflict-free, and have the maximum light efficiency output under the physical layout. Finally, the angle-intensity combinations of all points are integrated to generate a multi-dimensional irradiation parameter matrix covering spatial direction, energy output, and wavelength configuration, obtaining the multi-dimensional light source irradiation parameters.

[0135] In one embodiment, the calculation formula for the target irradiation angle is

[0136]

[0137]

[0138] The calculation formula for the target irradiation intensity is

[0139]

[0140]

[0141] Where is the target irradiation angle, is the incident angle of external light entering the target egg, is the incident angle of the current material in the target egg entering the next material, is the surface roughness, is the comprehensive amplitude reflection coefficient, is the refractive index of different materials in the target egg, is the thickness of different materials in the target egg, is the interface layer integration identifier of the transfer matrix method in the reflection calculation, TMM is the transfer matrix method, is the roughness correction factor; is the target irradiation intensity, is the incident light intensity, is the attenuation caused by absorption or scattering of each layer material of the target egg, is the absorption coefficient of different materials, is the actual path of photons changing with the incident angle in different materials, is the comprehensive amplitude transmission coefficient, is the interface layer integration identifier of the transfer matrix method in the transmission calculation.

[0142] In this embodiment, by selecting the point with the minimum eggshell roughness in each clear illumination area as the target illumination point, the interference of surface microstructures on light reflection and scattering is effectively avoided, and the stability and consistency of spectral signals are improved. By combining the illumination wavelength parameters and the information on the substances penetrated by the eggs, reflection analysis and transmission analysis are respectively carried out, which can accurately calculate the optimal illumination angle and the required illumination intensity, and realize the customized regulation of the light propagation path under complex curved surface structures. Finally, the illumination angle and intensity parameters are fused to construct a multi-dimensional light source illumination parameter that highly matches the target detection requirements, so as to maximize the detection signal-to-noise ratio while ensuring the excitation efficiency. The adaptive illumination optimization based on surface quality and material properties is realized, significantly enhancing the accuracy, sensitivity and robustness of egg spectral detection, and providing key light source control support for non-contact intelligent detection systems.

[0143] It should be understood that although the steps in the flowcharts involved in the above-described embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise clearly stated in this article, the execution of these steps is not strictly restricted by order, and these steps can be executed in other orders.

[0144] Based on the same inventive concept, an embodiment of the present application also provides an optical-spectrum-based egg quality identification device for implementing the above-mentioned optical-spectrum-based egg quality identification method. As Figure 9 shown, it includes: a feature recognition module 902, an illumination planning module 904, a spectral detection module 906, and a quality analysis module 908. The solution provided by this device to solve the problem is similar to the solution described in the above method. Therefore, the specific limitations in one or more embodiments of the optical-spectrum-based egg quality identification device provided below can refer to the limitations on an optical-spectrum-based egg quality identification method in the above text, and will not be repeated here.

[0145] In an exemplary embodiment, a computer device is provided. This computer device can be a server, and its internal structure diagram can be as Figure 10 shown. This computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Those skilled in the art can understand that Figure 10 the structure shown in is only a block diagram of some structures related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0146] In one embodiment, a computer device is further provided, which includes a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the steps in the above method embodiments are implemented.

[0147] In one embodiment, a computer-readable storage medium is provided, which stores a computer program. When the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0148] In one embodiment, a computer program product or a computer program is provided. The computer program product or the computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the steps in the above method embodiments.

[0149] Those of ordinary skill in the art can understand that all or part of the processes of implementing the above method embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it may include the processes of the above method embodiments.

[0150] The above embodiments merely represent several implementation manners of the present application. The description thereof is relatively specific and detailed, but it should not be construed as a limitation on the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.

Claims

1. An egg quality identification method based on optical spectrum, characterized in that, The method includes: Performing shell physical feature recognition on the multi-dimensional captured images of the target eggs to obtain the shell physical feature data of the target eggs; Removing the stain feature data in the shell physical feature data to obtain the shell contour feature data; Calculating the irradiation curvature range of the target eggs according to the light wavelength parameter and the shell color parameter of the target eggs; Selecting each light source irradiation area from the shell contour feature data according to the irradiation curvature range; Controlling the egg cleaning nozzle corresponding to the target eggs to clean each of the light source irradiation areas to obtain each clear irradiation area of the target eggs; For any one of the clear irradiation areas, selecting the point with the minimum shell roughness in the clear irradiation area as the target irradiation point; Performing reflection analysis on the light source irradiation angle according to the light wavelength parameter and the penetration substance information of the target eggs to obtain the target irradiation angle; And performing transmission analysis on the light source irradiation intensity according to the light wavelength parameter and the penetration substance information of the target eggs to obtain the target irradiation intensity; Fusing each of the target irradiation angles and each of the target irradiation intensities to obtain the multi-dimensional light source irradiation parameters; Controlling each modulation excitation source to perform time-resolved spectroscopy detection on the target eggs according to the multi-dimensional light source irradiation parameters to obtain each egg spectroscopy detection data; Evaluating the egg health characteristics of the target eggs according to each of the egg spectroscopy detection data to obtain the egg quality data of the target eggs.

2. The method according to claim 1, characterized in that, The evaluating the egg health characteristics of the target eggs according to each of the egg spectroscopy detection data to obtain the egg quality data of the target eggs includes: Performing time decay analysis on each of the egg spectroscopy detection data respectively to obtain each decay analysis feature vector; Analyzing the health characteristics of the target eggs according to each of the decay analysis feature vectors to obtain the egg health feature data; Predicting the healthy development trend of the egg health feature data to obtain the egg quality data.

3. The method according to claim 2, wherein The analyzing the health characteristics of the target eggs according to each of the decay analysis feature vectors to obtain the egg health feature data includes: Integrating each of the decay analysis feature vectors according to the time identifiers in each of the decay analysis feature vectors to obtain the decay data integration matrix; Performing multi-dimensional local feature extraction on the decay data integration matrix according to the health impact feature factors of the target eggs to obtain each local decay feature data; Performing dimensionality reduction fusion on each of the local decay feature data to obtain the decay feature supervector; Inputting the decay feature supervector into the health characteristic analysis model of the target eggs to obtain the egg health feature data.

4. The method according to claim 2, wherein The predicting the healthy development trend of the egg health feature data to obtain the egg quality data includes: Identifying the physical analysis data and the biological analysis data of the target eggs from the egg health feature data; Predicting the abnormal biological development of the target eggs according to the physical analysis data and the biological analysis data to obtain the biological evolution prediction data of the target eggs. Analyze the health trend of the target egg according to the biological evolution prediction data to obtain the egg quality data.

5. The method according to claim 1, wherein The calculation formula of the target irradiation angle is The calculation formula of the target irradiation intensity is Among them, is the target irradiation angle, is the incident angle of external light entering the target egg, is the incident angle when the current material in the target egg enters the next material, is the surface roughness, is the comprehensive amplitude reflection coefficient, is the refractive index of different materials in the target egg, is the thickness of different materials in the target egg, is the interface layer integration identifier of the transfer matrix method in the reflection calculation, TMM is the transfer matrix method, is the roughness correction factor; is the target irradiation intensity, is the incident light intensity, is the attenuation caused by absorption or scattering of each layer of materials in the target egg, is the absorption coefficient of different materials, is the actual path of photons changing with the incident angle in different materials, is the comprehensive amplitude transmission coefficient, is the interface layer integration identifier of the transfer matrix method in the transmission calculation.

6. An egg quality identification device based on optical spectrum, which is used to implement the steps of the method described in claim 1, and is characterized in that, The device includes: A feature recognition module, configured to identify the physical characteristics of the eggshell from the multi-dimensional captured image of the target egg to obtain the physical characteristic data of the eggshell of the target egg; An irradiation planning module, configured to remove the stain characteristic data in the physical characteristic data of the eggshell to obtain the eggshell contour characteristic data; Calculate the irradiation curvature range of the target egg according to the light wavelength parameter and the eggshell color parameter of the target egg; Select each light source irradiation area from the eggshell contour characteristic data according to the irradiation curvature range; Control the egg cleaning gun head corresponding to the target egg to clean each of the light source irradiation areas to obtain each clear irradiation area of the target egg; For any one of the clear irradiation areas, select the point with the minimum eggshell roughness in the clear irradiation area as the target irradiation point; Perform reflection analysis on the light source irradiation angle according to the light wavelength parameter and the penetration substance information of the target egg to obtain the target irradiation angle; And perform transmission analysis on the light source irradiation intensity according to the light wavelength parameter and the penetration substance information of the target egg to obtain the target irradiation intensity; Fuse each of the target irradiation angles and each of the target irradiation intensities to obtain multi-dimensional light source irradiation parameters; A spectral detection module, configured to control each modulation excitation source to perform time-resolved spectral detection on the target egg according to the multi-dimensional light source irradiation parameters to obtain each egg spectral detection data; A quality analysis module, configured to evaluate the egg health characteristics of the target egg according to each of the egg spectral detection data to obtain the egg quality data of the target egg.

7. A computer, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, the steps of the method according to any one of claims 1 to 5 are implemented.

Images