White-strip pig intelligent identification and automatic warehouse-in and warehouse-out management system

Abstract

The invention belongs to the technical field of informatization management, and relates to an intelligent identification and automatic warehouse-in and warehouse-out management system for white-strip pigs, and the system comprises a carcass collection module which is used for collecting the related data information of the carcass of the white-strip pigs; the model building module is used for building a three-dimensional standard carcass model with a space coordinate system and calculating the epidermis wrinkle degree and the fat thickness of the carcass of the white pig; the intelligent sorting module is used for setting the skin wrinkle degree of the white pig carcasses and the threshold value of carcass fat thickness deviation, and grading the white pig carcasses according to the data; the carcass identification module is used for identifying and recording the carcass of the white pig; the automatic warehouse-in and warehouse-out module is used for selecting different warehouse-in and warehouse-out and delivery modes according to the classification grades of the carcasses of the white-strip pigs; and the process optimization module is used for optimizing and improving the existing production process. The problem that a traditional white pig management mode is challenged in cost control and benefit improvement is solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of informatization management and relates to an intelligent identification and automated warehousing and outbound management system for dressed pigs. Background Art

[0002] In traditional dressed pig management and logistics processes, operations such as visual inspection, sorting, identification, warehousing and outbound, and distribution rely on manual labor. However, there are some inherent problems in this traditional visual inspection method. Manual inspection is easily affected by human factors, such as the experience of the inspector, fatigue level, and distraction of attention. These factors may make it difficult to ensure the accuracy and consistency of inspection results.

[0003] The quality and grade of dressed pig carcasses often rely on the subjective judgment of inspectors, lacking objective and unified standards, resulting in different evaluation results for dressed pigs of the same batch due to different inspectors. Manual sorting and identification require a large amount of time and effort and are prone to errors and omissions. The warehousing and outbound and distribution processes also lack intelligent and automated support and are difficult to meet the requirements of modern logistics for speed, accuracy, and traceability.

[0004] In view of the above problems, traditional dressed pig management methods face challenges in cost control and efficiency improvement. Due to the large number of manual participation links, high labor costs, and lack of intelligent support for inventory control and distribution strategies, it is difficult to reduce costs and improve efficiency. Summary of the Invention

[0005] To solve the above problems, the present invention provides an intelligent identification and automated warehousing and outbound management system for dressed pigs.

[0006] An intelligent identification and automated warehousing and outbound management system for dressed pigs includes the following modules:

[0007] A carcass acquisition module for acquiring relevant data information of dressed pig carcasses to provide basic data for subsequent model construction and intelligent sorting operations;

[0008] A model construction module for establishing a three-dimensional standard carcass model with a spatial coordinate system to calculate the epidermal fold degree and fat thickness of dressed pig carcasses;

[0009] An intelligent sorting module for setting thresholds for the deviation of the epidermal fold degree and carcass fat thickness of dressed pig carcasses and grading the dressed pig carcasses according to the data;

[0010] A carcass identification module for identifying and recording dressed pig carcasses for subsequent tracking and management;

[0011] The automated warehousing module selects different warehousing, distribution, and delivery modes according to the four grades of white-striped pig carcass classification.

[0012] The process optimization module is used to optimize and improve the existing production process to improve production efficiency and product quality.

[0013] Preferably, the carcass acquisition module includes several groups of high-resolution industrial cameras, ultrasonic detection equipment, and laser three-dimensional scanners.

[0014] Several groups of high-resolution industrial cameras are used to scan the surface morphology of the white-striped pig carcass and capture the characteristics of the epidermal wrinkles and color uniformity of the white-striped pig.

[0015] The ultrasonic detection equipment selects a reference point according to the spinal line of the white-striped pig carcass calibrated manually and measures the subcutaneous fat thickness in real time.

[0016] The laser three-dimensional scanner is used to scan the size of the white-striped pig carcass to form an original point cloud representing the white-striped pig carcass and provide spatial coordinate data for sorting.

[0017] Preferably, establishing a three-dimensional standard carcass model with a spatial coordinate system includes the following steps:

[0018] Taking the thoracolumbar junction of the white-striped pig carcass as the origin, a spatial coordinate system for the white-striped pig carcass is established. The spinal direction of the white-striped pig carcass is the X-axis, the transverse direction of the white-striped pig carcass is the Y-axis, and the direction from the back to the abdomen of the white-striped pig carcass is the Z-axis. The position data of the white-striped pig carcass collected by the laser three-dimensional scanner is input into the spatial coordinate system to form an original point cloud representing the white-striped pig carcass.

[0019] Combining the standard position data of the white-striped pig carcass recorded in history and inputting it into the spatial coordinate system to form an original point cloud representing the standard white-striped pig carcass, and using the original point cloud representing the standard white-striped pig carcass to construct a three-dimensional standard carcass model.

[0020] Preferably, calculating the epidermal wrinkle degree of the white-striped pig carcass includes the following steps:

[0021] Based on the analysis of surface curvature, the epidermal wrinkle degree δ of the white-striped pig carcass is calculated, satisfying the following formula:

[0022]

[0023] where δ represents the epidermal wrinkle degree; k max (p i ) represents the maximum principal curvature of point p i , and k min (p i ) represents the minimum principal curvature of point p i .ref (p i ) represents the average curvature of the corresponding area of the three-dimensional standard carcass model, combined with anatomical references; n represents the number of valid point clouds in the detection area.

[0024] Preferably, calculating the fat thickness of a white-striped pig carcass includes the following steps:

[0025] Based on the radial basis function (RBF), calculate the fat thickness field of the white-striped pig, which satisfies the following formula,

[0026]

[0027] where d(c) represents the predicted value of the fat thickness at position c of the white-striped pig, and c represents the three-dimensional coordinates (x, y, z) of the point to be predicted; c i represents the coordinates of the i-th reference point, which needs to be calibrated through ultrasonic actual measurement; m represents the number of selected reference points; w i represents the weight coefficient of the i-th radial basis function; represents the radial basis function, which satisfies the following formula,

[0028]

[0029] where r = ||c - c i || represents the Euclidean distance from the prediction point c to the i-th reference point c i ; δ represents the width coefficient of the Gaussian kernel, which is the smoothness control of the prediction result; the larger the value of δ, the smoother the prediction result;

[0030] The weight coefficient w i is solved by the following regularized least squares method,

[0031]

[0032] where m is the same as m in the fat thickness field formula of the white-striped pig; d(j) represents the predicted thickness of the j-th actual measurement point of the white-striped pig; d j represents the fat thickness measured by ultrasonic for the j-th actual measurement point; λ represents the regularization coefficient.

[0033] Preferably, setting the thresholds of the epidermal fold degree and the deviation of the carcass fat thickness of the white-striped pig carcass, and grading the white-striped pig carcass according to the data, includes the following steps:

[0034] Combined with the experimental data of the epidermal fold degree of the historical white-striped pig carcasses, pre-set the minimum threshold δ 1 and the maximum threshold δ 2 ;

[0035] Combined with the experimental data of the fat thickness of the dressed hog carcass in the historical record, preset the minimum error threshold ε 1 and the maximum error threshold ε 2 ;

[0036] For the dual-threshold combined determination of the epidermal fold degree and the fat thickness, set the grades of the dressed hog carcasses as special grade, first grade, second grade, and substandard grade;

[0037] When the epidermal fold degree is δ≤δ 1 , and the error between the predicted fat thickness and the measured fat thickness is |ε|≤ε 1 , the dressed hog meets both conditions and is of special grade;

[0038] When the epidermal fold degree is δ 1 <δ≤δ 2 , and the error between the predicted fat thickness and the measured fat thickness is ε 1 <|ε|≤ε 2 , if the dressed hog meets one of the conditions and the other is less than the minimum threshold, it is of first grade; if the dressed hog meets both conditions, it is of second grade;

[0039] When the epidermal fold degree is δ>δ 2 , and the error between the predicted fat thickness and the measured fat thickness is |ε|>ε 2 , if the dressed hog meets one of the conditions, it is of substandard grade.

[0040] Preferably, the carcass identification module includes an ink printing unit, a laser engraving unit, and a transportation track monitoring unit;

[0041] The ink printing unit is used to identify the judgment grade of the dressed hog carcass, and according to different judgment grades, select inks of different colors to print the "grade and batch number" identification;

[0042] Specifically, for the "special grade" dressed hogs classified by the intelligent sorting module, the ink printing unit prints the "Grade A" identification + batch number with "green" ink;

[0043] For the "first grade and second grade" dressed hogs classified by the intelligent sorting module, the ink printing unit prints the "Grade B 1 Grade, Grade B 2 Grade" identification + batch number with "yellow" ink;

[0044] For the "substandard grade" dressed hogs classified by the intelligent sorting module, the ink printing unit prints the "Grade C" identification + batch number with "red" ink;

[0045] The laser engraving unit is used to engrave the barcode label of the recorded information on the dressed hog carcass, and the judgment grade, warehousing information, outbound information, and shipping information of the dressed hog carcass will be entered into the corresponding barcode label;

[0046] The transportation trajectory monitoring unit is used to implant the UWB positioning chip into the subcutaneous layer of the carcass, and the implantation depth into the subcutaneous layer of the carcass is ≤ 1 mm. The UWB positioning chip can detect whether the white-striped pig carcass deviates from the preset route during the distribution process, and can quickly locate for illegal traceability.

[0047] Preferably, for the four grades of white-striped pig carcass classification, different inbound, outbound and distribution modes are selected, including the following steps:

[0048] For the "super grade" white-striped pigs, the scanning system in the cold storage recognizes the green "Grade A" logo + batch number printed with ink, and distributes the "super grade" white-striped pigs to the core shelf area with constant temperature and humidity. The core shelf area is equipped with independent temperature and humidity monitoring equipment;

[0049] The cold storage environment is forcibly set to -1°C ± 0.2°C, stored away from light, and the maximum stacking layer number is ≤ 2 layers;

[0050] If the inventory time of the "super grade" white-striped pigs exceeds 48 hours and they are not out of the warehouse, they will be automatically promoted to the highest outbound priority. During the outbound period, the "super grade" white-striped pigs only match high-end orders, are independently packaged with vacuum + ice coat, the transport vehicle is pre-cooled to -18°C and is equipped with a multi-band temperature recorder;

[0051] For the "first grade" white-striped pigs, the scanning system in the cold storage recognizes the yellow "Grade B" logo + batch number printed with ink. At this time, the "first grade" white-striped pigs share the main cold storage passage with the "super grade" white-striped pigs; 1 The "first grade" white-striped pigs are prohibited from being mixed with the "super grade" white-striped pigs, and the physical separation distance is ≥ 1.5 m, and the maximum stacking layer number is ≤ 4 layers;

[0052] During the outbound period, comprehensive scheduling is carried out according to "first in, first out = weight of 60% + order urgency = weight of 40%", and they are independently packaged with vacuum + ice coat. The transport vehicle is pre-cooled to -18°C and is equipped with a multi-band temperature recorder.

[0053] During the outbound period, comprehensive scheduling is carried out according to "first in, first out = weight of 60% + order urgency = weight of 40%", and they are independently packaged with vacuum + ice coat. The transport vehicle is pre-cooled to -18°C and is equipped with a multi-band temperature recorder.

[0054] Preferably, for the four grades of white-striped pig carcass classification, different inbound, outbound and distribution modes are selected, and the following steps are also included:

[0055] For the "second grade" white-striped pigs, the scanning system in the cold storage recognizes the yellow "Grade B" logo + batch number printed with ink, and distributes the "second grade" white-striped pigs to the standard shelf area without independent temperature and humidity monitoring equipment; 2 The "second grade" white-striped pigs are distributed to the standard shelf area without independent temperature and humidity monitoring equipment;

[0056] The maximum stacking layers of the standard shelves ≤ 6 layers. When the inventory of "secondary" white-striped pigs exceeds the limit, an alarm is triggered and it is prompted that "price cuts and promotions" are needed.

[0057] During the outbound period, it is mainly used to supplement the gaps in regular orders. Mixing with non-fresh products during distribution is allowed, but the temperature needs to be maintained ≤ 4°C.

[0058] For "substandard" white-striped pigs, the cold storage scanning system recognizes the red "Grade C" label + batch number printed in ink. At this time, the "substandard" white-striped pigs are allocated to the edge area of the cold storage with weaker cold air coverage.

[0059] The maximum stacking layers ≤ 8 layers, but the stacking stability needs to be inspected daily.

[0060] During the outbound period, it is preferentially matched with low-price bulk orders. Non-cold chain transportation using "insulated box + dry ice" is allowed during distribution, but the temperature needs to be maintained ≤ 8°C.

[0061] Preferably, it is used to optimize and improve the existing production process to improve production efficiency and product quality, including the following steps:

[0062] Extract corresponding samples from the white-striped pigs of the four grades for manual re-inspection. Compare whether the white-striped pigs judged by the three-dimensional standard carcass model meet the standards of this grade, and dynamically select the appropriate optimization mode according to the qualified rate of the manual re-inspection.

[0063] Combined with the historical record data and expert experience, according to the qualified rate Q of the manual re-inspection, set the minimum qualified rate threshold Q 1 and the maximum qualified rate threshold Q 2 ;

[0064] When Q < Q 1 , the production line is forced to stop for adjustment, and the three-dimensional standard carcass model needs to be retrained.

[0065] When Q 1 ≤ Q < Q 2 , automatically fine-tune the recognition parameters of the ultrasonic detection equipment, and adjust the camera frame rate and light intensity of the high-resolution industrial camera.

[0066] When Q ≥ Q 2 , continue to work while maintaining the current operating state.

[0067] To sum up, the present invention includes the following beneficial technical effects:

[0068] 1. Through automated and intelligent management, the labor cost is reduced; inventory can be precisely controlled, reducing inventory backlogs and waste, and lowering storage costs; the efficient inbound and outbound and distribution models reduce losses during transportation and distribution, further reducing costs.

[0069] 2. Intelligent sorting and identification improve the accuracy and efficiency of sorting, enabling white-striped pigs to quickly and accurately enter the corresponding sales channels, thereby enhancing sales effectiveness; the automated warehousing and distribution mode speeds up the logistics process, shortens the delivery cycle, improves customer satisfaction and loyalty, and thus increases market benefits.

[0070] 3. Intelligently adjust inventory levels and distribution strategies based on actual demand and historical data, optimizing resource allocation; through precise cost control and benefit analysis, the system can provide real-time cost-benefit ratio information for enterprises, helping them make more informed decisions; long-term intelligent and automated management helps enterprises establish a more stable, efficient, and sustainable production and logistics system, thereby enhancing long-term economic benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 It discloses a framework schematic diagram of an intelligent identification and automated warehousing and outbound management system for white-striped pigs. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0072] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

[0073] The following will Figure 1 make a detailed description of the present invention.

[0074] Referring to Figure 1 as shown, the present invention provides an intelligent identification and automated warehousing and outbound management system for white-striped pigs, including: a carcass collection module, a model construction module, an intelligent sorting module, a carcass identification module, an automated warehousing and outbound module, and a process optimization module.

[0075] The carcass collection module is used to collect relevant data information of white-striped pig carcasses, providing basic data for subsequent model construction and intelligent sorting operations.

[0076] Specifically, the carcass collection module includes several groups of high-resolution industrial cameras, ultrasonic detection equipment, and laser three-dimensional scanners;

[0077] Several groups of high-resolution industrial cameras are used to scan the surface morphology of white-striped pig carcasses, capturing features such as epidermal wrinkles and color uniformity of white-striped pigs;

[0078] The ultrasonic detection equipment selects reference points based on the spinal line of the white-striped pig carcass calibrated manually and measures the subcutaneous fat thickness in real time;

[0079] A laser three-dimensional scanner is used to scan the dimensions of a dressed pig carcass to form an original point cloud representing the dressed pig carcass, providing spatial coordinate data for sorting.

[0080] A model construction module establishes a three-dimensional standard carcass model with a spatial coordinate system, which is used to calculate the epidermal fold degree and fat thickness of the dressed pig carcass.

[0081] Specifically, taking the thoracolumbar junction of the dressed pig carcass as the origin, a spatial coordinate system for the dressed pig carcass is established. The spinal column direction of the dressed pig carcass is the X-axis, the transverse direction of the dressed pig carcass is the Y-axis, and the direction from the back to the abdomen of the dressed pig carcass is the Z-axis. The position data of the dressed pig carcass collected by the laser three-dimensional scanner is input into the spatial coordinate system to form an original point cloud representing the dressed pig carcass.

[0082] The standard position data of the dressed pig carcass in combination with historical records is input into the spatial coordinate system to form an original point cloud representing the standard dressed pig carcass, and a three-dimensional standard carcass model is constructed using the original point cloud representing the standard dressed pig carcass.

[0083] The collected original point cloud P = {p i} of the dressed pig carcass is aligned to the three-dimensional standard carcass model Q = {q i} through the Iterative Closest Point (ICP) algorithm, and the minimum error function is calculated.

[0084]

[0085] where R represents the rotation matrix (aligning the original point cloud of the dressed pig carcass to the three-dimensional standard carcass model), t represents the translation vector, which is the position compensation of the spatial coordinate system, and N represents the total number of matching point pairs between the original point cloud of the dressed pig carcass and the three-dimensional standard carcass model.

[0086] In one embodiment of the present invention, calculating the epidermal fold degree of the dressed pig carcass includes the following steps:

[0087] Based on the analysis of surface curvature, the epidermal fold degree δ of the dressed pig carcass is calculated, satisfying the following formula:

[0088]

[0089] where δ represents the epidermal fold degree; k max (p i ) represents the maximum principal curvature of point p i , and k min (p i ) represents the minimum principal curvature of point p i ; k ref (pi ) represents the average curvature value of the corresponding area of the three-dimensional standard carcass model, combined with the anatomical reference value; n represents the number of valid point clouds in the detection area.

[0090] Exemplarily, the bar code label encodes information of a white-striped pig with HX2305 - 7D, a mass of 75.3 Kg, a body length (X-axis) of 104.2 cm, and a ventral thickness (Z-axis) of 21.5 cm; the number of point clouds n = 1500 in the detection area of the back of this white-striped pig;

[0091] Assume that the point clouds in the back area satisfy k max = 0.085, k min = 0.024, k avg = 0.042;

[0092] Calculate the epidermal fold degree δ of the white-striped pig carcass, which satisfies the following formula,

[0093]

[0094] In one embodiment of the present invention, calculating the fat thickness of a white-striped pig carcass includes the following steps:

[0095] Based on the radial basis function (RBF), calculate the fat thickness field of the white-striped pig, which satisfies the following formula,

[0096]

[0097] where d(c) represents the predicted fat thickness value at position c of the white-striped pig, and c represents the three-dimensional coordinates (x, y, z) of the point to be predicted; c i represents the coordinates of the i-th reference point, which needs to be calibrated by ultrasonic actual measurement in advance (for example: evenly select 9 points along the backbone line of the white-striped pig); m represents the number of selected reference points; w i represents the weight coefficient of the i-th radial basis function; represents the radial basis function (using a Gaussian kernel function), which satisfies the following formula,

[0098]

[0099] where r = ||c - c i || represents the Euclidean distance from the prediction point c to the i-th reference point c i , and δ represents the width coefficient of the Gaussian kernel, which is the smoothness control of the prediction result; the larger the value of δ, the smoother the prediction result;

[0100] The weight coefficient w i is solved by the following regularized least squares method,

[0101]

[0102] Among them, m is the same as that in the formula of the fat thickness field of the dressed pig; d(j) represents the predicted thickness of the j-th measured point of the dressed pig's position; d j represents the fat thickness measured by ultrasonic waves at the j-th measured point; λ represents the regularization coefficient, which is used to prevent overfitting.

[0103] Exemplarily, for a dressed pig with the coding information HX2305-7D, the mass is 75.3 Kg, the body length (X-axis) is 104.2 cm, and the ventral thickness (Z-axis) is 21.5 cm; three reference points are evenly selected along the ridge line of the dressed pig carcass:

[0104] The coordinates of the first reference point C 1 (12.4, 0.0, 8.2), and the measured fat thickness d 1 = 18.5;

[0105] The coordinates of the second reference point C 2 (24.8, 0.0, 9.7), and the measured fat thickness d 2 = 22.1;

[0106] The coordinates of the third reference point C 3 (37.2, 0.0, 11.3), and the measured fat thickness d 3 = 24.8;

[0107] Set the width coefficient δ of the Gaussian kernel to 15 cm, and calculate Taking C 1 and C 2 as examples for calculation:

[0108]

[0109] Taking C 1 、C 2 、C 3 as examples, the complete kernel matrix:

[0110]

[0111] Set λ = 0.01. According to the regularized least squares method, the weight coefficient w i is solved through the following equation,

[0112]

[0113] Among them, represents the transpose self-multiplication matrix (3X3) of the kernel matrix; λI represents the regularized additional diagonal matrix (3X3, with the diagonal being λ); d represents C 1 、C 2 、C 3The measured fat thickness vector is d = [18.5, 22.1, 24.8] T , it can be seen that

[0114]

[0115] Adding the regularization term λI = 0.01I, it can be seen that

[0116]

[0117] Calculate it can be seen that

[0118]

[0119] Use the method of matrix inversion to solve

[0120] w = [w 1 , w 2 , w 3 = [19.2, 4.6, 2.1]

[0121] Assume that the three-dimensional coordinates of the point c to be predicted are c = (18.6, 0.0, 8.9). Calculate the distance from each reference point and substitute it into the calculation formula of the fat thickness field of the white-striped pig carcass

[0122]

[0123] d(c) = 19.2×0.871 + 4.6×0.866 + 2.1×0.204 = 20.3mm

[0124] The fat thickness measured by ultrasonic for the measured point is 19.8mm. Therefore, the error between the predicted fat thickness and the measured fat thickness is 0.5mm

[0125] The intelligent sorting module sets the thresholds for the epidermal fold degree and the deviation of the carcass fat thickness of the white-striped pig carcass, and classifies the white-striped pig carcass according to the data

[0126] Specifically, combining the experimental data of the epidermal fold degree of the white-striped pig carcass recorded in history, pre-set the minimum fold degree threshold δ 1 , the maximum fold degree threshold δ 2 ;

[0127] Combining the experimental data of the fat thickness of the white-striped pig carcass recorded in history, pre-set the minimum error threshold ε 1 , the maximum error threshold ε 2 ;

[0128] For the joint determination of the double thresholds of the epidermal fold degree and the fat thickness, set the grades of the white-striped pig carcass as special grade, first grade, second grade, and substandard grade

[0129] When the epidermal fold degree is δ ≤ δ 1 , the error between the predicted fat thickness and the measured fat thickness is |ε| ≤ ε 1 , and the white strip pig meets both conditions, which is a special grade;

[0130] When the epidermal fold degree is δ 1 < δ ≤ δ 2 , the error between the predicted fat thickness and the measured fat thickness is ε 1 < |ε| ≤ ε 2 , if the white strip pig meets one of the conditions and the other is less than the minimum threshold, it is a first grade; if the white strip pig meets both conditions, it is a second grade;

[0131] When the epidermal fold degree is δ > δ 2 , the error between the predicted fat thickness and the measured fat thickness is |ε| > ε 2 , if the white strip pig meets one of the conditions, it is a secondary grade.

[0132] Exemplarily, for a white strip pig with the coding information HX2308 - 5D, the epidermal fold degree of the carcass is δ = 0.02;

[0133] The detection point coordinates are (18.6, 0.0, 8.9), the predicted thickness is 20.8mm, the measured thickness is 21.3mm, and the deviation is - 0.5mm;

[0134] The detection point coordinates are (33.6, 0.0, 11.1), the predicted thickness is 24.7mm, the measured thickness is 23.9mm, and the deviation is 0.8mm;

[0135] So the average error is 0.65mm; the minimum threshold of the pre - set fold degree δ 1 = 0.015, the maximum threshold of the fold degree δ 2 = 0.025; the minimum threshold of the pre - set error ε 1 = 0.5mm, the maximum threshold of the error ε 2 = 1.0mm;

[0136] When the epidermal fold degree is 0.015 < 0.02 ≤ 0.025, the error between the predicted fat thickness and the measured fat thickness is 0.5 < |0.8| ≤ 1.0, and the white strip pig meets both conditions, which is a second grade.

[0137] The carcass identification module is used to identify and record the carcass of the white strip pig for subsequent tracking and management.

[0138] Specifically, the carcass identification module includes an ink printing unit, a laser engraving unit, and a transportation trajectory monitoring unit;

[0139] An ink printing unit, which is used to identify the judgment grade of a white-striped pig carcass, and select inks of different colors to print the "grade and batch number" identification according to different judgment grades;

[0140] Specifically, for the "super grade" white-striped pigs classified by the intelligent sorting module, the ink printing unit prints the "Grade A" identification + batch number with "green" ink;

[0141] For the "first grade and second grade" white-striped pigs classified by the intelligent sorting module, the ink printing unit prints the "B 1 Grade, B 2 Grade" identification + batch number with "yellow" ink;

[0142] For the "substandard grade" white-striped pigs classified by the intelligent sorting module, the ink printing unit prints the "Grade C" identification + batch number with "red" ink;

[0143] A laser engraving unit, which is used to engrave bar code labels for recording information on the white-striped pig carcass. The judgment grade, warehousing information, outbound information, and shipping information of the white-striped pig carcass will be entered into the corresponding bar code labels;

[0144] A transportation track monitoring unit, which is used to implant a UWB positioning chip into the subcutaneous layer of the carcass, and the implantation depth into the subcutaneous layer of the carcass ≤ 1 mm. The UWB positioning chip can detect whether the white-striped pig carcass deviates from the preset route during the distribution process, and can quickly locate for illegal traceability.

[0145] An automated warehousing and outbound module, which selects different warehousing, outbound, and distribution modes for the four grades of white-striped pig carcasses.

[0146] Specifically, for the "super grade" white-striped pigs, the scanning system in the cold storage recognizes the green "Grade A" identification + batch number printed with ink, and allocates the "super grade" white-striped pigs to the core shelf area with constant temperature and humidity. The core shelf area is equipped with independent temperature and humidity monitoring equipment;

[0147] The cold storage environment is forcibly set to -1°C ± 0.2°C, stored in the dark, and the maximum stacking layer ≤ 2 layers;

[0148] If the inventory time of the "super grade" white-striped pigs exceeds 48 hours and they have not been out of the warehouse, they will be automatically upgraded to the highest outbound priority. During the outbound period, the "super grade" white-striped pigs only match high-end orders (such as five-star hotels and premium supermarkets), are independently packaged with vacuum + ice coating, and the transport vehicle is pre-cooled to -18°C and equipped with a multi-band temperature recorder.

[0149] For the "first grade" white-striped pigs, the scanning system in the cold storage recognizes the yellow "B 1 Grade" identification + batch number printed with ink. At this time, the "first grade" white-striped pigs share the main cold storage passage with the "super grade" white-striped pigs;

[0150] "First-class" white strip pigs and "super-class" white strip pigs are prohibited from being mixed and stored, and the physical separation distance ≥ 1.5 m, and the maximum stacking layer number ≤ 4 layers;

[0151] During the outbound period, comprehensive scheduling is carried out according to "first in, first out = 60% weight + order urgency = 40% weight". Independent packaging of vacuum + ice coat is adopted, and the transport vehicle is pre-cooled to -18 °C and equipped with a multi-band temperature recorder.

[0152] For "second-class" white strip pigs, the cold storage scanning system identifies the yellow "B 2 class" logo + batch number printed with ink, and allocates the "second-class" white strip pigs to the standard shelf area without independent temperature and humidity monitoring equipment;

[0153] The maximum stacking layer number of the standard shelf ≤ 6 layers. When the inventory of "second-class" white strip pigs exceeds the limit, an alarm is triggered and it is prompted that "price cuts and promotions" are needed;

[0154] During the outbound period, it is mainly used to supplement the gaps in regular orders. Distribution is allowed to be mixed with non-fresh goods, but the temperature needs to be kept ≤ 4 °C.

[0155] Exemplarily, when the inventory of "second-class" white strip pigs exceeds the limit, an alarm is triggered and it is prompted that "price cuts and promotions" are needed, and the flash sale activity is automatically connected to the e-commerce platform to start.

[0156] For "substandard-class" white strip pigs, the cold storage scanning system identifies the red "C-class" logo + batch number printed with ink. At this time, the "substandard-class" white strip pigs are allocated to the cold storage edge area with weak cold air coverage;

[0157] The maximum stacking layer number ≤ 8 layers, but the stack stability needs to be inspected daily;

[0158] During the outbound period, it is preferentially matched with low-price bulk orders (for example: raw materials for food processing plants, animal feed for zoos). Distribution is allowed to use non-cold chain transportation of "insulated box + dry ice", but the temperature needs to be kept ≤ 8 °C.

[0159] The process optimization module is used to optimize and improve the existing production process to improve production efficiency and product quality.

[0160] For the four grades of white strip pigs, corresponding samples are extracted for manual re-inspection, and it is compared whether the white strip pigs judged by the three-dimensional standard carcass model meet the standards of this grade, and the appropriate optimization mode is dynamically selected through the qualified rate of manual re-inspection;

[0161] Combined with the historical record data and expert experience, according to the qualified rate Q of manual re-inspection, the minimum qualified rate threshold Q 1 and the maximum qualified rate threshold Q2 ;

[0162] When Q < Q 1 , the production line is forced to stop for adjustment, and the three-dimensional standard carcass model needs to be retrained;

[0163] When Q 1 ≤ Q < Q 2 , automatically fine-tune the recognition parameters of the ultrasonic detection equipment, and adjust the camera frame rate and light intensity of the high-resolution industrial camera;

[0164] When Q ≥ Q 2 , continue to work while maintaining the current operating state.

[0165] Define the interfaces and communication protocols between each module, clarify the interface specifications between each module, including: interface name, parameter type, return value, etc., and formulate corresponding communication protocols to ensure that data and information can be transferred correctly and efficiently between modules.

[0166] The above embodiments are only used to illustrate the technical solutions of the present invention; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A white-striped pig intelligent identification and automated in-and-out storage management system, characterized in that: Includes the following modules: The carcass collection module is used to collect relevant data information of white-striped pig carcasses and provide basic data for subsequent model construction and intelligent sorting operations; Model building module, which builds a three-dimensional standard carcass model with a spatial coordinate system to calculate the skin wrinkles and fat thickness of white striped pig carcasses; The intelligent sorting module sets the thresholds for the skin wrinkles and carcass fat thickness deviation of white-striped pig carcasses, and grades the white-striped pig carcasses based on the data; Carcass identification module, used to identify and record white-striped pig carcasses for subsequent tracking and management; The automated warehousing and delivery module selects different warehousing and delivery modes for the four grades of white-striped pig carcasses; The process optimization module is used to optimize and improve the existing production process to improve production efficiency and product quality.

2. The intelligent identification and automated storage and warehousing management system for white-striped pigs according to claim 1 is characterized by: Carcass collection module, including several sets of high-resolution industrial cameras, ultrasonic detection equipment, and laser 3D scanners; Several sets of high-resolution industrial cameras are used to scan the surface morphology of white-striped pig carcasses and capture the characteristics of white-striped pig skin wrinkles and color uniformity; Ultrasonic testing equipment selects reference points based on the manually calibrated ridgeline of the white-striped pig carcass and measures the thickness of subcutaneous fat in real time; The laser 3D scanner is used to scan the size of the white-striped pig carcass to form an original point cloud that represents the white-striped pig carcass and provide spatial coordinate data for sorting.

3. The intelligent identification and automated in-and-out storage management system for white-striped pigs according to claim 2 is characterized in that: Establishing a three-dimensional standard carcass model with a spatial coordinate system includes the following steps: Taking the thoracic and lumbar junction of the striped pig carcass as the origin, a spatial coordinate system about the striped pig carcass is established, the spine direction of the striped pig carcass is the X-axis, the transverse direction of the striped pig carcass is the Y-axis, and the direction from the back of the striped pig carcass to the abdomen is the Z-axis; the position data of the striped pig carcass collected by the laser three-dimensional scanner is input into the spatial coordinate system to form an original point cloud representing the striped pig carcass; The standard position data of the striped pig carcass combined with the historical records are input into the spatial coordinate system to form an original point cloud representing the standard striped pig carcass, and the original point cloud representing the standard striped pig carcass is used to construct a three-dimensional standard carcass model.

4. The intelligent identification and automated in-and-out storage management system for white-striped pigs according to claim 3 is characterized in that: Calculating the skin wrinkle degree of white striped pig carcass includes the following steps: Based on the analysis of surface curvature, the skin wrinkle degree δ of white striped pig carcass is calculated to meet the following formula: Among them, δ represents the degree of skin wrinkling; k max (p i ) represents point p i The maximum principal curvature, k min (p i ) represents point p i The minimum principal curvature of ref (p i ) represents the average curvature of the corresponding area of ​​the three-dimensional standard carcass model, combined with the anatomical reference; n represents the number of valid point clouds in the detection area.

5. The intelligent identification and automated in-and-out storage management system for white-striped pigs according to claim 4 is characterized in that: Calculating the fat thickness of a white pork carcass includes the following steps: Based on the radial basis function (RBF), the fat thickness field of white-striped pigs is calculated to meet the following formula: Where d(c) represents the predicted value of fat thickness at position c of the white striped pig, and c represents the three-dimensional coordinates (x, y, z) of the point to be predicted; c i represents the coordinates of the i-th reference point, which needs to be calibrated by ultrasonic measurement; m represents the number of selected reference points; w i represents the weight coefficient of the i-th radial basis function; represents the radial basis function, which satisfies the following formula: Where r = || cc i || indicates the predicted point c to the i-th reference point c i The Euclidean distance, δ represents the width coefficient of the Gaussian kernel, which controls the smoothness of the prediction result; the larger the value of δ, the smoother the prediction result; Weight coefficient w i Solved by the following regularized least squares method, Where m is the same as the m in the fat thickness field formula of white-striped pigs; d(j) represents the predicted thickness of the jth measured point of the white-striped pig; d j represents the fat thickness measured by ultrasound at the jth measuring point; λ represents the regularization coefficient.

6. The intelligent identification and automated in-and-out storage management system for white-striped pigs according to claim 5 is characterized in that: Setting the thresholds of the skin wrinkle degree and carcass fat thickness deviation of the white-striped pig carcass and grading the white-striped pig carcass according to the data includes the following steps: Combined with the historical experimental data of skin wrinkle degree of white striped pig carcasses, the minimum wrinkle degree threshold δ1 and the maximum wrinkle degree threshold δ2 are pre-set; Combined with the fat thickness experimental data of white-striped pig carcasses recorded in history, the minimum error threshold ε1 and the maximum error threshold ε2 are pre-set; Based on the dual thresholds of skin wrinkle and fat thickness, the grades of white-striped pig carcasses are set as special grade, first grade, second grade, and second grade. When the skin wrinkle degree is δ≤δ1, the error between the predicted fat thickness and the measured fat thickness is |ε|≤ε1, and the white-striped pig meets both conditions, that is, it is a special grade; When the skin wrinkle degree is δ1<δ≤δ2, and the error between the predicted fat thickness and the measured fat thickness is ε1<|ε|≤ε2, the white-striped pig meets one of the conditions and the other is less than the minimum threshold, which is the first level; the white-striped pig meets both conditions, which is the second level; When the skin wrinkle degree is δ>δ2, the error between the predicted fat thickness and the measured fat thickness is |ε|>ε2. If the white-striped pig meets one of the conditions, it is considered secondary.

7. The intelligent identification and automated storage and warehousing management system for white-striped pigs according to claim 6 is characterized by: Carcass identification module, including ink printing unit, laser engraving unit, and transportation track monitoring unit; The ink printing unit is used to identify the evaluation grade of the white-striped pig carcass, and select different colors of ink to print the "grade and batch number" logo according to different evaluation grades; Specifically, for the "special grade" white pork classified by the intelligent sorting module, the ink printing unit prints the "Grade A" logo + batch number using "green" ink; For the "first-grade and second-grade" carcasses classified by the intelligent sorting module, the ink printing unit prints the "B1 grade, B2 grade" logo + batch number using "yellow" ink; For the "secondary" carcasses classified by the intelligent sorting module, the ink printing unit prints the "C-grade" logo + batch number using "red" ink; The laser engraving unit is used to engrave barcode labels with recorded information on the carcasses of white-striped pigs. The evaluation grade, storage information, outbound information, and delivery information of the carcasses of white-striped pigs will be recorded in the corresponding barcode labels; The transport trajectory monitoring unit is used to implant the UWB positioning chip into the subcutaneous layer of the carcass. The depth of implantation into the subcutaneous layer of the carcass is ≤1mm. The UWB positioning chip can detect whether the white-striped pig carcass deviates from the preset route during distribution, and can quickly locate and trace violations.

8. The intelligent identification and automated in-and-out storage management system for white-striped pigs according to claim 7 is characterized in that: According to the four grades of white pork carcass classification, different storage and distribution modes are selected, including the following steps: For "special grade" carcass pork, the cold storage's scanning system recognizes the green "Grade A" logo + batch number printed in ink, and allocates the "special grade" carcass pork to the core shelf area with constant temperature and humidity. The core shelf area is equipped with independent temperature and humidity monitoring equipment; The cold storage environment is set to -1℃±0.2℃, stored away from light, and the maximum number of stacking layers is ≤2; If the "special grade" carcass pork has been in stock for more than 48 hours and has not been shipped out, it will automatically be upgraded to the highest shipping priority. During the shipping period, the "special grade" carcass pork will only be matched with high-end orders, and will be individually packaged with vacuum + ice coating. The transport vehicle will be pre-cooled to -18℃ and equipped with a multi-band temperature recorder. For "first-grade" carcass pigs, the cold storage scanning system recognizes the yellow "B1 grade" logo + batch number printed in ink. At this time, the "first-grade" carcass pigs and "special grade" carcass pigs share the main channel of the cold storage; "First-grade" carcass pigs and "special-grade" carcass pigs are not allowed to be mixed, and the physical separation distance is ≥1.5m, and the maximum number of stacking layers is ≤4; During the outbound process, comprehensive scheduling is carried out according to "First In, First Out = 60% weight + Order Urgency = 40% weight". Independent packaging with vacuum + ice coating is adopted, and the transport vehicle is pre-cooled to -18°C and equipped with a multi-band temperature recorder.

9. The intelligent identification and automated in-and-out storage management system for white-striped pigs according to claim 7, characterized in that: For the four grades of white-striped pig carcass classification, different inbound / outbound and distribution models are selected, and the following steps are also included: For "Grade 2" white-striped pigs, the cold storage scanning system recognizes the yellow "B2 Grade" label + batch number printed with ink, and allocates the "Grade 2" white-striped pigs to the standard shelf area without independent temperature and humidity monitoring equipment; The maximum stacking layers of the standard shelf ≤ 6 layers. When the inventory of "Grade 2" white-striped pigs exceeds the limit, an alarm is triggered and it is prompted that "price cuts for promotion" are needed; During the outbound process, it is mainly used to supplement the gaps in regular orders. Mixing with non-fresh food products is allowed during distribution, but the temperature needs to be maintained ≤ 4°C; For "Substandard" white-striped pigs, the cold storage scanning system recognizes the red "C Grade" label + batch number printed with ink. At this time, the "Substandard" white-striped pigs are allocated to the edge area of the cold storage with weak cold air coverage; The maximum stacking layers ≤ 8 layers, but the stack stability needs to be inspected daily; During the outbound process, it is preferentially matched with low-price bulk orders. Non-cold chain transportation with "insulated box + dry ice" is allowed during distribution, but the temperature needs to be maintained ≤ 8°C.

10. The intelligent identification and automated in-and-out storage management system for white-striped pigs according to claim 7, characterized in that: It is used to optimize and improve the existing production process to improve production efficiency and product quality, including the following steps: Extract corresponding samples from white-striped pigs of the four grades for manual re-inspection, compare whether the white-striped pigs judged by the three-dimensional standard carcass model meet the standards of this grade, and dynamically select the appropriate optimization mode according to the qualified rate of manual re-inspection; Combined with historical record data and expert experience, according to the qualified rate Q of manual re-inspection, set the minimum qualified rate threshold Q1 and the maximum qualified rate threshold Q2; When Q < Q1, the production line is forced to stop for adjustment, and the three-dimensional standard carcass model needs to be retrained; When Q1 ≤ Q < Q2, automatically fine-tune the recognition parameters of the ultrasonic detection equipment, and adjust the camera frame rate and light intensity of the high-resolution industrial camera; When Q ≥ Q2, continue to work in the current operating state.

Images