Egg quality identification method and device based on optical spectrum and computer

Abstract

The invention relates to an egg quality identification method and device based on an optical spectrum and a computer. The method comprises the following steps: performing eggshell physical feature recognition on a multi-dimensional shot image of a target egg to obtain eggshell physical feature data of the target egg; according to the eggshell physical characteristic data, performing light source space angle irradiation planning on the target egg to obtain a multi-dimensional light source irradiation parameter of the target egg; according to the multi-dimensional light source irradiation parameters, each modulation excitation source is controlled to carry out time-resolved spectrum detection on a target egg to obtain spectrum detection data of each egg; and according to the spectrum detection data of each egg, evaluating the egg health characteristics of the target egg to obtain egg quality data of the target egg. By adopting the method, the accuracy and efficiency of egg quality identification can be effectively improved, and the automation degree and the industrial application value in industrial production are further improved.

Description

Technical Field

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

[0002] In traditional technologies, a high-definition camera is used to collect the external image of an egg, and color, shape, cracks and other characteristic features are extracted through image preprocessing (such as denoising, segmentation, and feature extraction). Then, these features are input into a trained classification model (such as SVM, random forest, or convolutional neural network CNN) to achieve automatic discrimination of egg quality, such as identifying whether it is damaged, whether the surface is clean, or grading the size. However, the traditional technologies do 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: 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; 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; 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; According to each egg spectroscopy detection data, evaluating the egg health characteristics of the target egg to obtain egg quality data of the target egg.

[0005] In a second aspect, the present application further provides an egg quality identification device based on optical spectrum, including: 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; 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; A spectroscopy detection module, configured to control each modulation excitation source to perform time-resolved spectroscopy detection on the target egg according to the multi-dimensional light source irradiation parameters to obtain each egg spectroscopy detection data; A quality analysis module is used to evaluate the egg health characteristics of the target egg based on the egg spectrum detection data of each egg, and obtain the egg quality data of the target egg.

[0006] In a third aspect, the present application also provides a computer, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, any step of an egg quality identification method based on optical spectrum is implemented.

[0007] The above-mentioned egg quality identification method, device and computer based on optical spectrum can accurately identify the physical characteristic information of the eggshell of the target egg, including surface texture, shape structure, color distribution, etc., by collecting and analyzing multi-dimensional images of the target egg, so as to provide a basis for the subsequent illumination planning of the spatial angle of the light source, 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 spectrum detection on the target egg to obtain rich and high-precision spectrum response data; further through in-depth analysis of these spectrum detection data, the internal health characteristics of the egg, such as freshness, protein quality, whether there are 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 degree of automation and industrial application value in industrial production. Description of the Drawings

[0008] 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 to be used in the description of the embodiments or related technologies. Obviously, the following drawings 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.

[0009] Figure 1 It is an application environment diagram of an egg quality identification method based on optical spectrum in an embodiment; Figure 2 It is a flow diagram of an egg quality identification method based on optical spectrum in an embodiment; Figure 3 It is a flow diagram of a first method for obtaining egg quality data in an embodiment; Figure 4 It is a flow diagram of a method for obtaining egg health characteristic data in an embodiment; Figure 5 It is a flow diagram of a second method for obtaining egg quality data in an embodiment; Figure 6 It is a flow diagram of a method for obtaining multi-dimensional light source illumination parameters in an embodiment; Figure 7 Schematic flow chart of the method for obtaining the illumination area of each light source in the first embodiment; Figure 8 Schematic flow chart of the method for obtaining the illumination area of each light source in the second embodiment; Figure 9 Block diagram of the structure of an egg quality identification device based on optical spectrum in an embodiment; Figure 10 Internal structure diagram of a computer device in an embodiment. Detailed implementation manners

[0010] 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.

[0011] A method for identifying the quality of eggs based on optical spectrum provided by an embodiment of the present application can be applied to an application environment as shown in Figure 1 wherein, 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.

[0012] In an exemplary embodiment, as shown in Figure 2 a method for identifying the quality of eggs based on optical spectrum is provided. Taking the application of this method to the server in Figure 1 as an example, the following steps 202 to 208 are included. Among them: 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.

[0013] Among them, the multi-dimensional captured image can be a set of image data obtained by comprehensively collecting the target egg using imaging techniques with multiple angles, multiple spectra, and even multiple time points. It includes not only traditional visible light images, but also may cover dimension information such as infrared images, ultraviolet images, depth images, etc., for obtaining detailed features of the surface and local structures of the egg from multiple spatial dimensions and spectral dimensions.

[0014] Among them, the shell physical feature recognition can be the processing and analysis of the multi-dimensional image to extract the feature information at the physical level on the surface of the egg shell, such as color distribution, texture structure, shape contour, stain distribution, crack condition, spots and pores, etc.

[0015] Among them, the physical characteristic data of the eggshell can be the quantitative description results of the physical properties of the eggshell 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.

[0016] Specifically, use a multi-dimensional imaging system (such as multi-angle visible light imaging, near-infrared imaging, etc.) to take full-angle pictures of the target eggs to obtain their multi-view image data; then use image processing and analysis algorithms (such as edge detection, texture analysis, color distribution recognition, etc.) 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, so as to generate the physical characteristic data of the eggshell.

[0017] Step 204: According to the physical characteristic data of the eggshell, perform a light source spatial angle irradiation plan for the target eggs to obtain the multi-dimensional light source irradiation parameters of the target eggs.

[0018] Among them, the light source spatial angle irradiation plan can be to formulate a light source irradiation plan that is most suitable for the current egg structure according to the physical characteristic data of the eggshell, combined with three-dimensional modeling and optical simulation analysis.

[0019] 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.

[0020] Specifically, perform three-dimensional modeling according to the physical characteristic data of the eggshell, reconstruct the spatial geometric model of the target eggs, combine the texture distribution, color characteristics and possible defect areas on the eggshell surface, analyze the spectral response sensitivity of each area, and obtain a reasonable area for egg detection irradiation; on this basis, use the light propagation simulation algorithm to simulate and calculate the lighting effects under different incident angles, light source positions and lighting intensities, and evaluate the influence of different irradiation parameters on the spectral detection quality of eggs; select a set of optimal light source configuration plans according to the simulation results, and determine the spatial positions, irradiation directions, irradiation angles and light intensity distributions of each modulation excitation source and other parameters, so as to form the multi-dimensional light source irradiation parameters.

[0021] 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 eggs to obtain the spectral detection data of each egg.

[0022] Among them, the modulation excitation source can be a light source device with controllable light-emitting ability, such as a tunable 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 eggs to generate spectral responses.

[0023] Among them, time-resolved spectroscopy detection can be a spectroscopy acquisition technique that not only acquires spectral data at different wavelengths but also records the dynamic change process of the spectrum at different time points.

[0024] Among them, the egg spectral detection data can be the spectral response results obtained by a time-resolved spectroscopy detection system under specific irradiation conditions, usually including multiple spectral dimension information such as reflectivity, transmittance, absorptivity, or fluorescence intensity, reflecting the physical and chemical states of the egg under specific wavelength excitation.

[0025] Specifically, according to the multi-dimensional light source irradiation parameters, each modulation excitation source (such as LED light sources, lasers, etc. with different wavelengths) is controlled to start sequentially at the specified spatial position and irradiation angle, and synchronous excitation is performed according to the set time sequence to ensure multi-period and hierarchical light excitation of the target egg. During the excitation, the time-resolved spectroscopy detection element starts to collect the spectral response signals of the egg synchronously in real time, capturing spectral changes such as reflection, transmission, and fluorescence of the egg material at different time points, thereby constructing various egg spectral detection data covering the time dimension, spectral dimension, and spatial response dimension.

[0026] 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.

[0027] Among them, the egg health characteristics can be the index status of the egg itself at the physiological or quality level, including freshness, internal structural integrity, whether there are foreign objects or cracks, the ratio and uniformity of the egg white and yolk, water content, etc.

[0028] Among them, the egg quality data can be the comprehensive quantitative evaluation results obtained based on the evaluation of the egg health characteristics, usually presented in the form of quality data, scoring data, whether it is qualified, treatment suggestion data, etc., used to guide the grading, packaging, transportation, or elimination treatment of eggs, and support the industrialized egg quality management process.

[0029] Specifically, preprocess the spectral detection data of each egg, including operations such as noise filtering, baseline correction, spectral normalization, etc. Then use a feature extraction algorithm to extract key spectral feature parameters from the preprocessed spectral detection data of each egg, such as the reflectivity of specific bands, the position of absorption peaks, the change in fluorescence intensity, etc.; based on existing health characteristic evaluation models (such as support vector machines, neural networks, or discriminant analysis, etc.), match and discriminate the extracted spectral features with health indicators of the egg, such as freshness, protein content, yolk status, whether there are cracks or impurities, etc.; on this basis, combined with the set quality evaluation standard system, conduct quantitative analysis on the eggs to generate egg quality data, including egg health scores, egg quality data, and corresponding recommended processing data, etc.

[0030] In the above egg quality identification method based on optical spectrum, 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 light 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 health characteristics of the egg, such as freshness, protein quality, whether there are 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 degree of automation and industrial application value in industrial production.

[0031] In an exemplary embodiment, as Figure 3 shown, according to the spectral detection data of each egg, evaluate the egg health characteristics of the target egg to obtain the egg quality data of the target egg, including steps 302 to 306. Among them: Step 302, perform time decay analysis on the spectral detection data of each egg respectively to obtain each decay analysis feature vector.

[0032] Among them, time decay analysis can be a technology for modeling and analyzing the process of the response signal of the egg 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 the egg to light stimulation, thereby revealing the activity, stability, and structural integrity of the egg material.

[0033] Among them, the decay analysis feature vector can be a vector set composed of key parameters extracted from time decay analysis, used to quantify the time response characteristics of the egg 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.

[0034] Specifically, time series analysis is performed on the obtained egg spectral detection data, with particular attention paid to the decay behavior of the spectral intensity over time under different wavelength excitations. By establishing a time decay curve and using fitting algorithms (such as exponential decay models, polynomial regression, etc.) to model each set of spectral data, key parameters describing the change speed, amplitude, and stability of the spectral response are extracted, such as the initial intensity, half-life time, decay constant, etc. Furthermore, a decay analysis feature vector corresponding to each detection channel is constructed, and the decay analysis feature vector can be used to characterize the dynamic response characteristics of the egg material to spectral stimuli.

[0035] Step 304: Analyze the health characteristics of the target egg based on each decay analysis feature vector to obtain egg health characteristic data.

[0036] Among them, the health characteristics can be internal indicators reflecting the overall quality and edible safety of the egg, including freshness, physical structure states of the egg white and yolk, cracks or impurities, internal moisture content, etc.

[0037] Among them, the egg health characteristic data can be a set of quantitative information that can specifically describe the health status of the egg generated based on the evaluation of the decay analysis feature vector, including indicators such as freshness data, structural integrity index, internal abnormality probability, etc., usually represented in the form of structured data.

[0038] Specifically, each decay analysis feature vector is input into a pre-trained health characteristic evaluation model. The health characteristic evaluation model establishes a mapping relationship between the decay characteristics and the internal state of the egg based on machine learning methods (such as support vector machines, random forests, or deep neural networks), performs classification or regression analysis on each input set of decay analysis feature vectors, and identifies physiological indicators such as the freshness level, structural integrity, and states of the egg white and yolk corresponding to them in the historical samples; according to the analysis results, multiple health-related quantitative indicators are extracted, such as spectral stability data, crack potential data, freshness data, etc., and these indicators are structurally integrated into a dataset reflecting the current internal health status of the egg, that is, the egg health characteristic data.

[0039] Step 306: Predict the healthy development trend of the egg health characteristic data to obtain egg quality data.

[0040] 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 change speed of the egg quality during storage or transportation, whether it will accelerate spoilage or remain stable.

[0041] Specifically, the egg health characteristic data is sorted according to the time series structure, and combined with the known egg storage rules and environmental parameters (such as temperature, humidity, light, etc.) to construct the input features of the prediction model. Using time series prediction algorithms, such as ARIMA model, LSTM neural network or ensemble learning method, the change trends of various health indicators (such as fresh data, structurally intact data, etc.) in the future period are modeled and predicted to evaluate the possible quality decline, spoilage risk or structural degradation over time; the predicted health change trend data is concatenated with the current characteristic state to output the egg quality data.

[0042] In this embodiment, by performing time decay analysis on the egg spectral detection data and extracting dynamic response feature vectors, the spectral change law of the egg at different time points can be more accurately reflected, and the perception ability of its internal state change can be improved; subsequently, based on these decay feature vectors, health characteristic analysis is carried out to realize the 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, the quality evolution of the egg during storage and transportation can be predicted in advance, 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 having good practical application value and industrial promotion prospects.

[0043] In an exemplary embodiment, as Figure 4 shown, according to each decay analysis feature vector, the health characteristics of the target egg are analyzed to obtain egg health characteristic data, including steps 402 to 408. Among them: Step 402, according to the time identifiers in each decay analysis feature vector, integrate each decay analysis feature vector to obtain a decay data integration matrix.

[0044] Among them, the decay data integration matrix can be a multi-dimensional data structure formed by uniformly integrating multiple decay analysis feature vectors according to the time sequence and feature dimension, and is used to comprehensively represent the spectral response changes of the egg at different time points and different wavelength bands during the spectral detection process.

[0045] Specifically, extract the time identifier information from each decay analysis feature vector, perform normalization arrangement and alignment on all vectors in chronological order, and then stack the data with the same time label according to the dimension to construct a multi-dimensional feature matrix reflecting the evolution of the decay process over time - that is, the decay data integration matrix, which still comprehensively retains the response characteristics of the target egg under different spectral conditions and different time points.

[0046] Step 404: Based on the health impact characteristic factors of the target eggs, perform multi-dimensional local feature extraction on the attenuation data integration matrix to obtain local attenuation feature data for each part.

[0047] Among them, the health impact characteristic factors can be key spectra or time attenuation characteristics that are representative and judgmentally significant for the health status of eggs, such as attenuation rate, instantaneous amplitude fluctuation, energy concentration region, peak shift, etc.

[0048] Among them, multi-dimensional local feature extraction can be a process of extracting fine-grained features with local representativeness for specific time windows and spectral dimension regions in the attenuation data integration matrix.

[0049] Among them, local attenuation feature data can be a structured feature set extracted from the attenuation matrix, reflecting the dynamic response characteristics within a specific time period or spectral interval. It describes information such as signal intensity changes, fluctuation frequencies, and attenuation patterns in the local area, and is an important basis for measuring whether there are abnormalities in the internal microscopic state of eggs.

[0050] Specifically, based on 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 specific local regions in the attenuation data integration matrix, and extract the fine-grained attenuation characteristics shown by each region 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 region, and refine key indicators that can reflect changes in health status, such as local energy distribution, mutation point location, or signal instability, etc.; these local attenuation feature data extracted from different regions.

[0051] Step 406: Perform dimensionality reduction and fusion on the local attenuation feature data for each part to obtain an attenuation feature supervector.

[0052] 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.

[0053] 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 throughout the detection period.

[0054] Specifically, normalize 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, removing redundant information and retaining the principal component features that contribute most to health discrimination; through methods such as feature concatenation, weighted averaging, or feature attention mechanism, fuse the dimensionality-reduced local attenuation feature data to integrate and form a unified and powerful low-dimensional vector, obtaining an attenuation feature supervector representing the overall health status of the eggs.

[0055] Step 408, input the attenuation feature supervector into the health characteristic analysis model of the target egg to obtain egg health characteristic data.

[0056] 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.

[0057] 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, identifying the corresponding health data in the training samples, and analyzing situations such as structurally abnormal data, quality deterioration data, freshness data, etc.; the egg health characteristic analysis model converts the analysis results into structured egg health characteristic data.

[0058] In this embodiment, by introducing time identifiers to integrate each attenuation analysis feature vector, a high-temporal-precision attenuation data integration matrix is constructed, retaining the complete information of the egg spectral signal changing over time; further combining health impact feature factors to perform multi-dimensional local feature extraction on this matrix, effectively capturing small but key spatio-temporal signal features, improving the sensitivity and precision of health status recognition; then performing dimensionality reduction and fusion 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, using the health characteristic analysis model to intelligently analyze the supervector and output accurate and quantifiable egg health characteristic data. It 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.

[0059] In an exemplary embodiment, such as Figure 5As shown, the healthy development trend of the egg health characteristic data is predicted to obtain the egg quality data, including steps 502 to 506. Among them: Step 502: Identify the physical analysis data and biological analysis data of the target egg from the egg health characteristic data.

[0060] Among them, physical analysis data can be quantitative information extracted from the external structure or surface characteristics of eggs, usually derived from physical perception methods such as image processing, structured light measurement or spectral reflectance characteristics.

[0061] Among them, bioanalytical data can be quantitative indicators obtained through time-resolved spectroscopy, fluorescence response, etc., which are used to reflect the internal physiological and chemical state of eggs. These data cover freshness, protein degradation, moisture content, internal air chamber changes, microbial activity, etc., and can dynamically reveal the potential quality changes of eggs during storage or transportation.

[0062] Specifically, various features of the egg health feature data are classified and identified through the feature labels, data source attributes and historical sample correspondences; data reflecting the external structural state of the egg (such as eggshell integrity, surface roughness, crack index, spectral reflectance characteristics, etc.) are extracted as physical analysis data based on the physical sources and perception methods of the classified and identified features. These are usually static structural indicators from image analysis or spectral signals; at the same time, indicators representing changes in the internal composition of the egg, physiological activity, micro-metabolism, etc. (such as freshness scores, moisture content trends, protein decay signals, volatile substance responses, etc.) are identified from the remaining features and divided into biological analysis data. Most of these features come from dynamic spectral analysis and time response characteristics.

[0063] 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.

[0064] 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.

[0065] 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.

[0066] Specifically, first identify the key features highly correlated with quality changes in the bioanalysis data, such as the protein degradation rate, the moisture evaporation trend, the air chamber expansion rate, and the dynamic changes of volatile components, etc.; then input these time-related biological characteristics and physical analysis data into the constructed bioabnormal development prediction model. The bioabnormal development prediction model is constructed based on time series algorithms (such as LSTM, GRU) or multivariate regression models, and is used to simulate the biological change process of eggs (such as the change in egg quality caused by gas exchange inside and outside the eggs, etc.) under the current physical analysis data and these time-related biological characteristics under specific storage conditions (such as temperature, humidity); the bioabnormal development prediction model predicts the change trends of these key indicators of the target eggs in the future time period by learning the evolution trajectories of historical sample data, and identifies development signs such as whether there is a potential sharp decline in freshness data, internal spoilage, or abnormal microbial activity, etc.; finally, output the above prediction results in a structured time series form to form the biological evolution prediction data of the target eggs.

[0067] Step 506, analyze the health trend of the target eggs based on the biological evolution prediction data to obtain egg quality data.

[0068] Among them, the health trend can be the development direction and change range of the overall health status of the eggs in the future for a period of time, usually obtained by comprehensive analysis of the current health characteristic data and the biological evolution prediction data.

[0069] Specifically, compare and analyze the biological evolution prediction data with the current health characteristic data to identify the change trends of key indicators (such as freshness data, moisture retention rate, internal structure integrity, etc.) in the time dimension; and based on the preset health development assessment model or threshold rules, 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 time, risk change data in the future for a period of time, etc.; then combine the storage environment parameters (such as temperature, humidity) involved in the prediction process for multi-scenario deduction to correct the health trend evolution path of the eggs under different logistics or storage conditions; finally, generate a set of egg quality data including the current quality grade, future edibility estimation, potential risk warnings and recommended treatment suggestions, etc.

[0070] In this embodiment, by accurately identifying 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 egg is achieved, which helps to more finely grasp the quality composition elements of the egg; 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 structural variation, and give early warnings of quality risks; finally, through the dynamic trend analysis of the biological evolution prediction data, egg quality data covering the current state and future change directions 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.

[0071] 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 egg to obtain multi-dimensional light source irradiation parameters of the target egg, including steps 602 to 606. Among them: Step 602, according to the eggshell physical characteristic data, analyze the light source irradiation range of the target egg to obtain each light source irradiation area of the target egg.

[0072] 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.

[0073] Among them, the light source irradiation area can be a 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, can produce 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.

[0074] Specifically, use three-dimensional reconstruction technology to construct the egg surface space model based on the eggshell physical characteristic data, clarify its physical attributes such as shape contour, surface curvature, texture direction, and color distribution; further use the light propagation simulation algorithm, combined with the optical characteristics such as reflectivity and absorptivity of the egg surface, to simulate and analyze the path and response of light irradiated on the egg surface at different angles and directions; according to the simulation results, identify which areas are prone to shadow occlusion, reflection interference, or information loss, and eliminate them, and preferentially retain the areas with good spectral response, flat surface, and stable characteristics as the light source action targets; according to the screening results of the areas, divide the egg surface into multiple independent light source irradiation areas.

[0075] Step 604: Control the egg cleaning nozzle corresponding to the target egg to clean each light source irradiation area, and obtain each clear irradiation area of the target egg.

[0076] Among them, the egg cleaning nozzle can be an automated equipment component used to precisely clean specific areas on the egg surface, and is usually integrated into the preprocessing module of the detection system.

[0077] Among them, the clear irradiation area can be an irradiation area whose surface reaches a clean, stain-free, and light-interference-free state after directional cleaning by the egg cleaning nozzle.

[0078] Specifically, due to the storage conditions of the eggs, it is not possible to clean the entire egg comprehensively. Therefore, the three-dimensional coordinates, size, and surface state information of each irradiation area are transmitted to the cleaning control module, and appropriate cleaning parameters, including nozzle type, cleaning angle, water pressure intensity, and duration, are intelligently matched according to the area size of the irradiation area, the type of surface attachments (such as dust, water stains, and stains), and the degree of interference they may cause to spectral reflection. Then, the control system activates the corresponding egg cleaning nozzle to perform fixed-point, directional, and high-precision cleaning on 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 a vision or reflectivity sensor, and secondary compensation processing is performed on areas that are not cleaned thoroughly. Finally, when all irradiation areas meet the conditions of surface cleanliness, no residue, controllable water film, and light penetration, they are marked as clear irradiation areas, providing a clean and stable optical response surface for subsequent high-quality light angle and intensity planning.

[0079] Step 606: Perform light source irradiation angle analysis and light source irradiation intensity analysis on each clear irradiation area respectively to obtain the multi-dimensional light source irradiation parameters of the target egg.

[0080] Among them, the light source irradiation angle analysis can be to perform light simulation and reflection simulation on each clear irradiation area to evaluate the effect of light at different incident angles, so as to 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.

[0081] 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.

[0082] Specifically, surface state sampling is performed on each clear illumination area to obtain its current reflection characteristics, surface normal direction, and local curvature information. Based on the egg surface model and optical properties, illumination simulation and reverse ray tracing algorithms are used to perform multi-angle simulation illumination analysis on each clear illumination area to evaluate the reflectivity, scattering pattern, and signal response intensity of light at different incident angles. Through comparative analysis, the optimal illumination angle is determined for each area, that is, the incident direction that can maximize the excitation of the spectral signal and minimize the specular reflection interference. Then, according to the spectral response sensitivity and target detection depth of each clear illumination area, combined with historical illumination experience and physical models, the required light source intensity is matched and optimized, including illumination power, modulation frequency, and wavelength range. The optimal angle parameters and illumination intensity parameters of each clear illumination area are summarized to generate a complete multi-dimensional light source illumination parameter set.

[0083] In this embodiment, by analyzing the light source illumination range of the eggshell physical characteristic data, the key areas on the egg surface that are most suitable for spectral detection can be accurately identified, avoiding illumination interference caused by abnormal curvature or surface defects. Further, by controlling the egg cleaning nozzle to perform fixed-point cleaning on each illumination area, surface stains and reflection interference objects are effectively removed, ensuring that the illumination area has ideal cleanliness and optical response conditions. Subsequently, the illumination angle and intensity of the cleaned area are analyzed, and the optimal illumination parameters are accurately calculated based on the surface structure and material penetration characteristics, realizing personalized and high-precision light source configuration. This ensures the stability of spectral excitation and the accuracy of detection data, significantly improving the reliability of subsequent spectral acquisition and egg quality analysis, and providing strong technical support for realizing intelligent, non-contact, and highly sensitive egg detection.

[0084] In an exemplary embodiment, as Figure 7 shown, according to the eggshell physical characteristic data, a light source illumination range analysis is performed on the target egg to obtain the light source illumination areas of the target egg, including steps 702 to 706. Among them: Step 702, the stain characteristic data in the eggshell physical characteristic data is removed to obtain the eggshell contour characteristic data.

[0085] Among them, the stain characteristic data can be the characteristic information related to the attachments on the egg surface identified in the eggshell image or spectral information, including dust, water stains, oil spots, foreign objects, etc.

[0086] Among them, the eggshell contour characteristic data can be the real eggshell geometric structure information extracted from the image or sensing information after removing the stains, mainly including the edge contour, surface curvature, normal direction, geometric continuity, and local texture distribution of the eggshell.

[0087] 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.

[0088] 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.

[0089] 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 egg.

[0090] 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, etc. Eggshells of different colors have different absorption and reflection capabilities for light of different wavelengths.

[0091] 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.

[0092] Specifically, after obtaining the characteristic data of the eggshell contour, the illumination wavelength parameters corresponding to the current detection task are called to clarify the wavelength range of the light source used and its reflection and absorption characteristics on different materials and color surfaces; then, according to the color parameters of the eggshell surface, combined with the spectral response curve of each color at a specific wavelength, the reflectivity and penetration ability of each color area to the illumination wave are analyzed. The color response characteristics are integrated with the curvature information of the eggshell contour to simulate the energy distribution and reflection efficiency when light is incident at different angles to areas with different curvature radii, and identify which curvature ranges can produce stable and usable spectral signals under the current light source conditions. Based on these simulation results, a set of numerical boundaries are established, and a set of illumination curvature ranges are output to screen those eggshell surface areas that can ensure good light incidence and facilitate subsequent signal acquisition.

[0093] Step 706, selecting the illumination areas of each light source from the eggshell contour feature data according to the illumination curvature range.

[0094] Specifically, after calculating the irradiation curvature range, this range is used as a screening condition and applied to the previously obtained eggshell contour feature data. By matching the local curvature values of each surface area, candidate areas that meet the irradiation curvature requirements are identified. Further, geometric visibility analysis is performed on these candidate areas to evaluate whether there are problems such as occlusion, shadow, or excessive incident angle under the light source layout of the detection system, and areas where light cannot reach or the light effect is poor are excluded; and the remaining areas are evaluated for optical performance. Based on the surface smoothness and the matching degree between the normal direction and the light source direction, it is determined whether an ideal reflection or transmission signal can be obtained under the preset wavelength conditions. All areas that pass the screening and evaluation are marked as formal light source irradiation areas, and each area is assigned accurate spatial coordinates, size information, and a candidate illumination angle range.

[0095] In this embodiment, by clearing the stain feature data in the eggshell physical feature data, surface interference factors that affect the accuracy of optical detection are effectively eliminated, improving the clarity and reliability of subsequent image and spectral analysis; after further extracting the eggshell contour feature data, combined with the illumination wavelength parameter and the eggshell color parameter, the best irradiation curvature range can be accurately calculated according to the response characteristics of different materials and colors to light of a specific wavelength; finally, the most suitable light source irradiation area for spectral excitation is selected 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 area. An automated and high-precision irradiation plan from feature extraction, condition matching to area screening is realized, significantly improving the accuracy and adaptability of egg spectral detection, and providing solid data and physical support for the intelligent detection system.

[0096] In an exemplary embodiment, as Figure 8 shown, light source irradiation angle analysis and light source irradiation intensity analysis are respectively performed on each clear irradiation area to obtain multi-dimensional light source irradiation parameters of the target egg, including steps 802 to 808. Among them: 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.

[0097] 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 surface height changes (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.

[0098] 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 characteristics such as a flat surface, small reflection interference, and stable signal response, and is the basic coordinate for spectral excitation and energy control calculation.

[0099] Specifically, high-resolution imaging data or laser point cloud data is called to obtain the surface texture and topography information of each tiny point in each area. Then, a roughness calculation algorithm (such as the Ra or RMS method based on height deviation) is used to evaluate the roughness of all points one by one, generating a roughness distribution map for each area. Further, in each area, the point with the minimum roughness value is located in the roughness distribution map, that is, the flattest and smoothest position on the surface. This point has the best reflection stability and the least scattering interference. This point is confirmed as the target irradiation point of the clear irradiation area, and its key information such as spatial coordinates and surface normal direction is recorded.

[0100] 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.

[0101] 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 of each layer, absorption coefficient, scattering coefficient, medium composition, etc.

[0102] Among them, the reflection analysis can be a process of simulating and evaluating the reflection intensity, directivity, and effectiveness of light when it irradiates the target point at different angles.

[0103] Among them, the target irradiation angle can be the optimal light incident direction obtained after reflection analysis, usually represented by the pitch angle and 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.

[0104] Specifically, for each area, based on the spatial coordinates and surface normal direction of the target irradiation point, combined with the preset light wavelength parameter 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 plan for this irradiation point.

[0105] Step 806: According to the light wavelength parameter and the penetration material information of the target egg, perform a transmission analysis on the light source irradiation intensity to obtain the target irradiation intensity.

[0106] 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 eggshells based on the wavelength of light, the irradiation angle, and the penetration material information.

[0107] 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.

[0108] Specifically, based on the spatial positions of the target irradiation angle and the target irradiation point as constraints, according to the preset light wavelength parameters and the information of the penetration material of the egg (such as the absorption coefficient, scattering coefficient, refractive index, etc. of the eggshell material), an optical transmission model (such as the Lambert-Beer law) is called to simulate the energy attenuation process of light penetration in the penetration material information, and the transmission threshold capable of generating an effective spectral response in the internal detection area under different light intensity conditions is calculated; further, the minimum or appropriate light energy required to ensure penetration to the required depth and excitation of a stable detection signal under the target irradiation angle is deduced, so as to determine the optimal light source output parameters, including power density, modulation frequency, wavelength bandwidth, etc. These parameters are output as the target irradiation intensity corresponding to this point.

[0109] Step 808, fuse each target irradiation angle and each target irradiation intensity to obtain multi-dimensional light source irradiation parameters.

[0110] Specifically, after the target irradiation angle and the target irradiation intensity corresponding to each clear irradiation area are respectively used as a set of data for structured arrangement as the light source layout, and then according to the spatial distribution and light demand of each point, the light source layout strategy is coordinated and optimized to ensure that the angle and intensity combinations in different areas are feasible, 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, and multi-dimensional light source irradiation parameters are obtained.

[0111] In one embodiment, the calculation formula for the target irradiation angle is The calculation formula for 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 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, and 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.

[0112] In this embodiment, by selecting the point with the smallest eggshell roughness in each clear irradiation area as the target irradiation 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 of substances penetrated by the egg, reflection analysis and transmission analysis are respectively carried out, and the optimal irradiation angle and the required illumination intensity can be accurately calculated, so as to realize the customized regulation of the light propagation path under the complex curved surface structure; finally, the irradiation angle and intensity parameters are fused to construct a multi-dimensional light source irradiation parameter that highly matches the target detection requirements, thereby maximizing the detection signal-to-noise ratio while ensuring the excitation efficiency. The adaptive illumination optimization based on surface quality and material properties is realized, and the accuracy, sensitivity and robustness of egg spectral detection are significantly enhanced, providing key light source control support for the non-contact intelligent detection system.

[0113] 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 there is a clear indication in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders.

[0114] Based on the same inventive concept, the 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 9As shown, it includes a feature recognition module 902, an irradiation planning module 904, a spectral detection module 906, and a quality analysis module 908. The implementation solution provided by this device to solve the problem is similar to the implementation solution described in the above method. Therefore, the specific limitations in one or more embodiments of the optical spectrum-based egg quality recognition device provided below can refer to the limitations of an optical spectrum-based egg quality recognition method in the above text, which will not be elaborated here.

[0115] 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 the figure is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this 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.

[0116] In an embodiment, a computer device is further provided, including 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.

[0117] In an embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0118] In an embodiment, a computer program product or a computer program is provided. This computer program product or 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 implements the steps in the above method embodiments.

[0119] Those of ordinary skill in the art can understand that all or part of the processes in 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, and when the computer program is executed, it can include the processes of the above method embodiments.

[0120] The above-described 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 to 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 fall within 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 comprises: Performing eggshell physical feature recognition on the multi-dimensional captured image of the target egg to obtain eggshell physical feature data of the target egg; According to the eggshell physical characteristic data, a light source spatial angle irradiation planning is performed on the target egg to obtain a multi-dimensional light source irradiation parameter of the target egg; According to the irradiation parameters of the multi-dimensional light source, each modulated excitation source is controlled to perform time-resolved spectral detection on the target egg to obtain spectral detection data of each egg; According to the spectral detection data of each egg, the egg health characteristics of the target egg are evaluated to obtain the egg quality data of the target egg.

2. The method according to claim 1, characterized in that The step of evaluating the egg health characteristics of the target eggs according to the egg spectrum detection data to obtain the egg quality data of the target eggs includes: Performing time decay analysis on each egg spectrum detection data to obtain each decay analysis feature vector; Analyze the health characteristics of the target eggs according to each of the attenuation analysis feature vectors to obtain egg health characteristic data; The health development trend of the egg health characteristic data is predicted to obtain the egg quality data.

3. The method according to claim 2, characterized in that The step of analyzing the health characteristics of the target eggs according to the attenuation analysis feature vectors to obtain egg health feature data includes: Integrate the attenuation analysis feature vectors according to the time identifiers in the attenuation analysis feature vectors to obtain an attenuation data integration matrix; According to the health impact characteristic factors of the target eggs, multi-dimensional local feature extraction is performed on the attenuation data integration matrix to obtain local attenuation characteristic data; Performing dimensionality reduction fusion on each of the local attenuation feature data to obtain an attenuation feature supervector; The attenuation characteristic supervector is input into the health characteristic analysis model of the target egg to obtain the egg health characteristic data.

4. The method according to claim 2, characterized in that: The method of predicting the health development trend of the egg health characteristic data to obtain the egg quality data includes: Identify the physical analysis data and biological analysis data of the target egg from the egg health characteristic data; According to the physical analysis data and the biological analysis data, predicting the abnormal biological development of the target egg, and obtaining biological evolution prediction data of the target egg; The health trend of the target eggs is analyzed according to the biological evolution prediction data to obtain the egg quality data.

5. The method according to claim 1, characterized in that The method of performing light source spatial angle irradiation planning on the target egg according to the eggshell physical characteristic data to obtain multi-dimensional light source irradiation parameters of the target egg includes: According to the eggshell physical characteristic data, analyzing the light source irradiation range of the target egg to obtain each light source irradiation area of ​​the target egg; 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; The light source illumination angle analysis and the light source illumination intensity analysis are respectively performed on each of the clear illumination areas to obtain the multi-dimensional light source illumination parameters of the target egg.

6. The method according to claim 5, characterized in that The step of analyzing the light source irradiation range of the target egg according to the eggshell physical characteristic data to obtain the irradiation areas of each light source of the target egg comprises: Clearing the stain feature data in the eggshell physical feature data to obtain eggshell contour feature data; Calculating the irradiation curvature range of the target egg according to the illumination wavelength parameter and the eggshell color parameter of the target egg; According to the illumination curvature range, each light source illumination area is selected from the eggshell contour feature data.

7. The method according to claim 6, characterized in that The light source irradiation angle analysis and light source irradiation intensity analysis are performed on each of the clear irradiation areas to obtain the multi-dimensional light source irradiation parameters of the target egg, including: For any of the clear irradiation areas, a point with the smallest eggshell roughness in the clear irradiation area is selected as a target irradiation point; According to the illumination wavelength parameter and the penetrating material information of the target egg, a reflection analysis is performed on the illumination angle of the light source to obtain a target illumination angle; And, according to the illumination wavelength parameter and the penetrating material information of the target egg, the illumination intensity of the light source is subjected to transmission analysis to obtain the target illumination intensity; The illumination angles of the targets and the illumination intensities of the targets are integrated to obtain the illumination parameters of the multi-dimensional light source.

8. The method according to claim 7, characterized in that The calculation formula of the target illumination angle is: The calculation formula of the target irradiation intensity is: in, is the target irradiation angle, is the incident angle of the external object into the target egg, is the incident angle of the current material into the next material in the target egg, 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, It is the interface layer integration mark of the transfer matrix method in reflection calculation. TMM is the transfer matrix method. is the roughness correction factor; is the target irradiation intensity, is the incident light intensity, The attenuation caused by absorption or scattering of each layer of the target egg material is is the absorption coefficient of different materials, is the actual path of photons in different materials as the incident angle changes, is the comprehensive amplitude transmission coefficient, Integrated markers for interface layers in the transfer matrix method in transmission calculations.

9. An egg quality identification device based on optical spectrum, characterized in that: The device comprises: A feature recognition module is used to recognize the physical features of the eggshell of the target egg in a multi-dimensional captured image to obtain the physical feature data of the eggshell of the target egg; An irradiation planning module, used to plan the spatial angle irradiation of the target egg with a light source according to the eggshell physical characteristic data, and obtain the multi-dimensional light source irradiation parameters of the target egg; A spectral detection module, used to control each modulated excitation source to perform time-resolved spectral detection on the target egg according to the irradiation parameters of the multi-dimensional light source, so as to obtain spectral detection data of each egg; The quality analysis module is used to evaluate the egg health characteristics of the target eggs according to the spectral detection data of each egg, so as to obtain the egg quality data of the target eggs.

10. A computer comprising a memory and a processor, wherein the memory stores 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 8 are implemented.

Images