CN116235799A - Live pig breeding weight real-time monitoring system based on image recognition - Google Patents
Live pig breeding weight real-time monitoring system based on image recognition Download PDFInfo
- Publication number
- CN116235799A CN116235799A CN202310192317.9A CN202310192317A CN116235799A CN 116235799 A CN116235799 A CN 116235799A CN 202310192317 A CN202310192317 A CN 202310192317A CN 116235799 A CN116235799 A CN 116235799A
- Authority
- CN
- China
- Prior art keywords
- weight
- value
- parameter
- growth
- state
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 142
- 238000009395 breeding Methods 0.000 title claims abstract description 20
- 230000001488 breeding effect Effects 0.000 title claims abstract description 20
- 241000282887 Suidae Species 0.000 claims abstract description 68
- 238000012545 processing Methods 0.000 claims abstract description 55
- 238000000034 method Methods 0.000 claims abstract description 38
- 238000003860 storage Methods 0.000 claims abstract description 19
- 238000005303 weighing Methods 0.000 claims abstract description 9
- 238000004891 communication Methods 0.000 claims abstract description 4
- 238000004458 analytical method Methods 0.000 claims description 33
- 230000002159 abnormal effect Effects 0.000 claims description 25
- 238000004364 calculation method Methods 0.000 claims description 9
- 230000037396 body weight Effects 0.000 claims description 6
- 238000012098 association analyses Methods 0.000 claims description 4
- 238000012935 Averaging Methods 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 238000010219 correlation analysis Methods 0.000 claims description 3
- 238000001514 detection method Methods 0.000 claims description 3
- 238000005211 surface analysis Methods 0.000 claims description 3
- 238000009313 farming Methods 0.000 claims 1
- 230000003938 response to stress Effects 0.000 abstract description 6
- 238000005516 engineering process Methods 0.000 abstract description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 238000007792 addition Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 235000013305 food Nutrition 0.000 description 1
- 235000013372 meat Nutrition 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- A—HUMAN NECESSITIES
- A01—AGRICULTURE; FORESTRY; ANIMAL HUSBANDRY; HUNTING; TRAPPING; FISHING
- A01K—ANIMAL HUSBANDRY; CARE OF BIRDS, FISHES, INSECTS; FISHING; REARING OR BREEDING ANIMALS, NOT OTHERWISE PROVIDED FOR; NEW BREEDS OF ANIMALS
- A01K29/00—Other apparatus for animal husbandry
-
- A—HUMAN NECESSITIES
- A01—AGRICULTURE; FORESTRY; ANIMAL HUSBANDRY; HUNTING; TRAPPING; FISHING
- A01K—ANIMAL HUSBANDRY; CARE OF BIRDS, FISHES, INSECTS; FISHING; REARING OR BREEDING ANIMALS, NOT OTHERWISE PROVIDED FOR; NEW BREEDS OF ANIMALS
- A01K29/00—Other apparatus for animal husbandry
- A01K29/005—Monitoring or measuring activity, e.g. detecting heat or mating
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10028—Range image; Depth image; 3D point clouds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30181—Earth observation
- G06T2207/30188—Vegetation; Agriculture
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30232—Surveillance
Abstract
The invention belongs to the field of live pig breeding, relates to a data processing technology, and is used for solving the problem that the death rate of a pig farm is increased due to the fact that a live pig weight acquisition mode in the prior art is manual weighing and stress response is brought to the live pig, in particular to a live pig breeding weight real-time monitoring system based on image recognition, which comprises a real-time monitoring platform, wherein the real-time monitoring platform is in communication connection with an image acquisition module, a data processing module, a model generation module, a state monitoring module and a storage module, and the image acquisition module is used for acquiring point cloud images of the live pig: the three-dimensional point cloud data of the live pigs are sent to a data processing module by the selected collected live pigs and the collection environment; the method is used for acquiring the point cloud image of the live pig and obtaining the three-dimensional point cloud data, and further processing the three-dimensional point cloud data through the data processing module to obtain the growth parameters of the live pig, so that the stress response of the live pig caused by manually weighing the weight of the live pig is avoided.
Description
Technical Field
The invention belongs to the field of live pig breeding, relates to a data processing technology, and particularly relates to a live pig breeding weight real-time monitoring system based on image recognition.
Background
The pig raising industry is an important industry in China's agriculture, plays an important role in guaranteeing the safety supply of meat foods, the pig raising industry in China is changed from the traditional pig raising industry to the modern pig raising industry, the raising mode, the regional layout, the production mode and the production capacity are all changed remarkably, the average daily gain is an important basis for judging the growth speed of pigs, and the average daily gain can only be obtained by means of a time-consuming and labor-consuming manual weighing mode at present, so that great stress response is brought to pigs, the death rate of pig farms is increased, and therefore, a breeder cannot acquire weight data of the pigs in time in the growth process of the pigs, and the pigs cannot be fed empirically and cannot be fed accurately;
aiming at the technical problems, the application provides a solution.
Disclosure of Invention
The invention aims to provide a live pig breeding weight real-time monitoring system based on image recognition, which is used for solving the problem that the death rate of a pig farm is increased due to stress response to live pigs because the live pig weight acquisition mode in the prior art is manual weighing.
The technical problems to be solved by the invention are as follows: how to provide a live pig breeding weight real-time monitoring system based on image recognition, which can automatically acquire live pig weight data.
The aim of the invention can be achieved by the following technical scheme:
the live pig breeding weight real-time monitoring system based on image recognition comprises a real-time monitoring platform, wherein the real-time monitoring platform is in communication connection with an image acquisition module, a data processing module, a model generation module, a state monitoring module and a storage module;
the image acquisition module is used for acquiring point cloud images of live pigs: scanning the live pigs in the selected collected live pigs and the collecting environment through AzureKinect to obtain three-dimensional point cloud data of the live pigs, and sending the three-dimensional point cloud data of the live pigs to a data processing module;
the data processing module is used for processing three-dimensional point cloud data of live pigs to obtain growth parameters i of the live pigs, i=1, 2, …, n and n are positive integers, a plurality of live pigs are selected as specimen objects, the growth parameters i of the specimen objects are subjected to association analysis, the growth parameters i are marked as irrelevant parameters i or association parameters i, and the association parameters i of the specimen objects are sent to the model generating module;
the model generation module is used for generating a weight estimation model through the associated parameters of the specimen object; the weight of the live pigs is monitored and analyzed through the weight estimation model, a weight output value is generated and output when the estimated and analyzed result meets the requirement, and an estimated abnormal signal is sent to the state monitoring module through the real-time monitoring platform when the estimated and analyzed result does not meet the requirement;
the state monitoring module is used for monitoring and analyzing the growth state of the monitored object.
As a preferred embodiment of the present invention, the specific process of performing the correlation analysis on the growth parameter i of the specimen object includes: the method comprises the steps of collecting a weight value of a specimen object and a numerical value of a growth parameter i, weighing the weight value of the specimen object, setting a weight range of the specimen object, dividing a maximum value and a minimum value of the weight range of the specimen object into a plurality of individual weight sections, marking the numerical value of the growth parameter i of the specimen object as SZix when the weight value of the specimen object is a minimum boundary value of the weight section, marking the numerical value of the growth parameter i of the specimen object as SZid when the weight value of the specimen object is a maximum boundary value of the weight section, marking the absolute value of a difference value between SZid and SZix as a change value BHi of the growth parameter i, acquiring a change threshold value BHmin through a storage module, comparing the change value BHi of the growth parameter i with the change threshold value BHmin, and marking the growth parameter as an irrelevant parameter i or a related parameter i through a comparison result.
As a preferred embodiment of the present invention, the specific process of comparing the variation value BHi of the growth parameter i with the variation threshold BHmin includes: if the variation value BHi is larger than or equal to the variation threshold value BHmin, marking the growth parameter i as an influence parameter i corresponding to the weight interval; if the variation value BHi is smaller than the variation threshold value BHmin, marking the growth parameter i as a random parameter i corresponding to the weight interval; marking the number ratio of the times that the growth parameter i is marked as an influence parameter in the weight range to the number of weight intervals as an influence coefficient, acquiring an influence threshold value through a storage module, and comparing the influence coefficient with the influence threshold value: if the influence coefficient is smaller than the influence threshold, marking the corresponding growth parameter i as an irrelevant parameter i of the specimen object; and if the influence coefficient is greater than or equal to the influence threshold, marking the corresponding growth parameter i as the associated parameter i of the specimen object.
As a preferred embodiment of the present invention, the specific process of generating the weight estimation model by the model generation module through the associated parameter i of the specimen object includes: summing all the SZix of the sample objects in the weight interval to obtain a low average value DJi of the associated parameter i, summing all the SZid of the sample objects in the weight interval to obtain a high average value GJi of the associated parameter i, marking the difference value between the maximum boundary value and the minimum boundary value of the weight interval as a body difference value TC, obtaining an influence coefficient YXi of the associated parameter i through the high average value GJi, the low average value DJi and the body difference value TCi, and generating a weight estimation model through the influence coefficient YXi of all the associated parameter i.
As a preferred embodiment of the invention, the specific process of the weight estimation model for carrying out weight monitoring analysis on the weight of the live pigs comprises the following steps: marking a live pig subjected to weight monitoring analysis as a monitoring object, acquiring and processing a point cloud image of the monitoring object to obtain a numerical value of an associated parameter i of the monitoring object, marking the numerical value as JCi, and obtaining a weight estimated value TZi corresponding to the associated parameter i through a formula TZi= | JCi-DJi | YXi + BJx, wherein BJx is a minimum boundary value corresponding to a weight interval; and establishing a weight set of weight estimation values TZi corresponding to all the associated parameters i, performing variance calculation on the weight set to obtain an estimation deviation coefficient, acquiring an estimation deviation threshold value through a storage module, comparing the estimation deviation coefficient with the estimation deviation threshold value, and judging whether an estimation analysis result meets the requirement or not through a comparison result.
As a preferred embodiment of the present invention, the specific process of comparing the estimated deviation coefficient with the estimated deviation threshold value includes: if the estimated deviation coefficient is smaller than the estimated deviation threshold, judging that the estimated analysis result meets the requirement, summing the weight estimated values TZi of all the associated parameters i, averaging to obtain a weight output value, and outputting the weight output value to a real-time monitoring platform; if the estimated deviation coefficient is larger than or equal to the estimated deviation threshold, judging that the estimated analysis result does not meet the requirement, sending an estimated abnormal signal to a real-time monitoring platform by the weight estimation model, and sending the estimated abnormal signal to a state monitoring module after the estimated abnormal signal is received by the real-time monitoring platform.
As a preferred embodiment of the present invention, the specific process of the state monitoring module for monitoring and analyzing the growth state of the monitored object includes: the method comprises the steps of performing image shooting on a monitoring object to obtain a monitoring image, amplifying the monitoring image into a pixel grid image, performing gray level conversion to obtain gray level values of the pixel grid, summing the gray level values of all the pixel grids, taking an average value to obtain color data SZ of the monitoring object, and performing body surface analysis on the monitoring object: the method comprises the steps that a gray threshold value is obtained through a storage module, pixel grids with the gray value larger than the gray threshold value are marked as spot grids, a pixel grid area formed by mutually contacted spot grids is marked as a spot area, the number of spot areas and the total number of the pixel grids in the spot area are obtained and respectively marked as BS and XS, and a state coefficient ZT of a monitoring object is obtained through numerical calculation of SZ, BS and XS; the state threshold value ZTmin is obtained through the storage module, the state coefficient ZT is compared with the state threshold value ZTmin, and whether the growth state of the live pigs is normal or not is judged according to a comparison result.
As a preferred embodiment of the present invention, the specific process of comparing the state coefficient ZT with the state threshold ZTmin includes: if the state coefficient ZT is smaller than the state threshold ZTmin, judging that the growth state of the monitored object is abnormal, sending an abnormal processing signal to a real-time monitoring platform by a state monitoring module, and sending the abnormal processing signal to a mobile phone terminal of a manager after the real-time monitoring platform receives the abnormal processing signal; if the state coefficient ZT is larger than or equal to the state threshold ZTmin, the growth state of the monitoring object is judged to be normal, the state detection module sends a data updating signal to the real-time monitoring platform, the real-time monitoring platform receives the data updating signal and then sends the data updating signal to the data processing module, and the data processing module acquires new live pig growth parameters for processing after receiving the data updating signal and updates a data source of the weight estimation model.
The working method of the live pig breeding weight real-time monitoring system based on image recognition comprises the following steps:
step one: and (3) collecting point cloud images of live pigs: the method comprises the steps of selecting a collected live pig and a collection environment, and scanning the live pig in the collection environment through AzureKinect to obtain three-dimensional point cloud data of the live pig;
step two: processing three-dimensional point cloud data of live pigs to obtain growth parameters i, i=1, 2, …, n and n of the live pigs, selecting a plurality of live pigs as specimen objects, collecting and processing weight values of the specimen objects and numerical values of the growth parameters i, and marking the growth parameters as irrelevant parameters or related parameters;
step three: body weight monitoring analysis was performed on live pig body weight: marking a live pig subjected to weight monitoring analysis as a monitoring object, acquiring and processing a point cloud image of the monitoring object to obtain a weight estimated value TZi and an estimated deviation coefficient corresponding to the related parameter i, and judging whether an estimated analysis result meets the requirement or not through the estimated deviation coefficient;
step four: when the estimated analysis result of the monitoring object does not meet the requirement, the growth state of the monitoring object is monitored and analyzed to obtain a state coefficient ZT of the monitoring object, and the reason that the estimated result of the weight does not meet the requirement is judged according to the value of the state coefficient ZT.
The invention has the following beneficial effects:
1. the method comprises the steps that point cloud images of live pigs can be acquired through an image acquisition module, three-dimensional point cloud data are obtained, the three-dimensional point cloud data are processed through a data processing module to obtain growth parameters of the live pigs, correlation analysis is carried out on the growth parameters to obtain correlation parameters which have an influence relationship with the weight change of the live pigs, the weight of the live pigs is estimated through monitoring of the correlation parameters, and stress response of the live pigs caused by manual weighing of the weight of the live pigs is avoided;
2. the method comprises the steps that the model generation module can analyze the associated parameters of a specimen object to obtain a weight estimation model, the weight range is decomposed into a plurality of individual weight intervals, the change state of the associated parameters in the weight intervals is subjected to linear simulation to obtain the influence coefficients of the associated parameters, and data support is provided for weight estimation analysis of the weight estimation model;
3. the weight of the live pigs can be monitored and analyzed through the weight estimation model, a plurality of individual weight estimated values are obtained through collecting and calculating each associated parameter, and the variance calculation is carried out on all the weight estimated values to obtain an estimated deviation coefficient, so that the accuracy of a calculation result is fed back according to the estimated deviation coefficient, and the accuracy of a weight output value is ensured;
4. the growth state of the monitored object can be monitored and analyzed through the state monitoring module, the state coefficient is obtained through comprehensive analysis of a plurality of parameters of the monitored image, the growth state of the live pigs is fed back through the numerical value of the state coefficient, the model data source is updated when the growth state is normal, and the abnormality processing is carried out when the growth state is abnormal.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings can be obtained according to these drawings without inventive effort to a person skilled in the art.
FIG. 1 is a system block diagram of a first embodiment of the present invention;
fig. 2 is a flowchart of a method according to a second embodiment of the invention.
Detailed Description
The technical solutions of the present invention will be clearly and completely described in connection with the embodiments, and it is obvious that the described embodiments are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Example 1
As shown in FIG. 1, the live pig breeding weight real-time monitoring system based on image recognition comprises a real-time monitoring platform, wherein the real-time monitoring platform is in communication connection with an image acquisition module, a data processing module, a model generation module, a state monitoring module and a storage module.
The image acquisition module is used for acquiring point cloud images of live pigs: and scanning the live pigs in the selected collected live pigs and the collecting environment through AzureKinect to obtain three-dimensional point cloud data of the live pigs, and sending the three-dimensional point cloud data of the live pigs to a data processing module.
The data processing module processes three-dimensional point cloud data of the live pigs to obtain growth parameters i, i=1, 2, …, n and n of the live pigs, an automatic control software platform such as a double-fortune TwainCAT 3 and Bei Jialai Automation studio and other related controllers are adopted, and an autonomously developed automatic weighting algorithm file is embedded into a depth image acquisition device through the controllers to realize data acquisition of the growth parameters i; selecting a plurality of live pigs as specimen objects, and carrying out association analysis on a growth parameter i of the specimen objects: the method comprises the steps of collecting a weight value of a specimen object and a numerical value of a growth parameter i, weighing the weight value of the specimen object, setting a weight range of the specimen object, dividing a maximum value and a minimum value of the weight range of the specimen object into a plurality of individual weight sections, marking the numerical value of the growth parameter i of the specimen object as SZix when the weight value of the specimen object is a minimum boundary value of the weight section, marking the numerical value of the growth parameter i of the specimen object as SZid when the weight value of the specimen object is a maximum boundary value of the weight section, marking the absolute value of a difference value between SZid and SZix as a change value BHi of the growth parameter i, acquiring a change threshold BHmin through a storage module, and comparing the change value BHi of the growth parameter i with the change threshold BHmin: if the variation value BHi is larger than or equal to the variation threshold value BHmin, marking the growth parameter i as an influence parameter corresponding to the weight interval; if the variation value BHi is smaller than the variation threshold value BHmin, marking the growth parameter i as a random parameter i corresponding to the weight interval; marking the number ratio of the times that the growth parameter i is marked as an influence parameter in the weight range to the number of weight intervals as an influence coefficient, acquiring an influence threshold value through a storage module, and comparing the influence coefficient with the influence threshold value: if the influence coefficient is smaller than the influence threshold, marking the corresponding growth parameter i as an irrelevant parameter i of the specimen object; if the influence coefficient is greater than or equal to the influence threshold, marking the corresponding growth parameter i as the associated parameter i of the specimen object; the method comprises the steps of sending a relevant parameter i of a specimen object to a model generation module; the method comprises the steps of collecting point cloud images of live pigs, obtaining three-dimensional point cloud data, further processing the three-dimensional point cloud data through a data processing module to obtain growth parameters of the live pigs, performing association analysis on the growth parameters to obtain association parameters which have an influence relationship with weight change of the live pigs, further estimating the weight of the live pigs through monitoring the association parameters, and avoiding live pig stress response caused by manual weighing of the weight of the live pigs.
The model generation module is used for generating a weight estimation model through the associated parameters i of the specimen object: summing all the SZix of the sample objects in the weight interval to obtain a low average value DJi of the associated parameter i, summing all the SZid of the sample objects in the weight interval to obtain a high average value GJi of the associated parameter i, marking the difference value between the maximum boundary value and the minimum boundary value of the weight interval as a body difference value TC, obtaining an influence coefficient YXi of the associated parameter i through a formula YXi = TC/(| GJi-DJi |), and generating a weight estimation model through the influence coefficient YXi of all the associated parameter i; analyzing the related parameters of the specimen object to obtain a weight estimation model, decomposing the weight range into a plurality of individual weight intervals, and further performing linear simulation on the change states of the related parameters in the weight intervals to obtain the influence coefficients of the related parameters, so as to provide data support for weight estimation analysis of the weight estimation model.
The weight estimation model is used for carrying out weight monitoring analysis on the weight of the live pigs: marking a live pig subjected to weight monitoring analysis as a monitoring object, acquiring and processing a point cloud image of the monitoring object to obtain a numerical value of an associated parameter i of the monitoring object, marking the numerical value as JCi, and obtaining a weight estimated value TZi corresponding to the associated parameter i through a formula TZi= | JCi-DJi | YXi + BJx, wherein BJx is a minimum boundary value corresponding to a weight interval; establishing a weight set of weight estimated values TZi corresponding to all the associated parameters i, performing variance calculation on the weight set to obtain an estimated deviation coefficient, acquiring an estimated deviation threshold value through a storage module, and comparing the estimated deviation coefficient with the estimated deviation threshold value: if the estimated deviation coefficient is smaller than the estimated deviation threshold, judging that the estimated analysis result meets the requirement, summing the weight estimated values TZi of all the associated parameters i, averaging to obtain a weight output value, and outputting the weight output value to a real-time monitoring platform; if the estimated deviation coefficient is larger than or equal to the estimated deviation threshold, judging that the estimated analysis result does not meet the requirement, sending an estimated abnormal signal to a real-time monitoring platform by the weight estimation model, and sending the estimated abnormal signal to a state monitoring module after the estimated abnormal signal is received by the real-time monitoring platform; and (3) carrying out weight monitoring analysis on the weight of the live pigs, acquiring and calculating each associated parameter to obtain a plurality of estimated weight values, and carrying out variance calculation on all the estimated weight values to obtain an estimated deviation coefficient, so that the accuracy of a calculation result is fed back according to the estimated deviation coefficient, and the accuracy of a weight output value is ensured.
The state monitoring module is used for monitoring and analyzing the growth state of the monitored object: the method comprises the steps of performing image shooting on a monitoring object to obtain a monitoring image, amplifying the monitoring image into a pixel grid image, performing gray level conversion to obtain gray level values of the pixel grid, summing the gray level values of all the pixel grids, taking an average value to obtain color data SZ of the monitoring object, and performing body surface analysis on the monitoring object: the method comprises the steps of obtaining a gray threshold through a storage module, marking pixel grids with the gray value larger than the gray threshold as spot grids, marking a pixel grid area formed by mutually contacted spot grids as spot areas, obtaining the number of the spot areas and the total number of the pixel grids in the spot areas, respectively marking the pixel grids as BS and XS, obtaining a state coefficient ZT of a monitoring object through a formula ZT=α1SZ/(α2xBS+α3xXS), wherein the state coefficient is a numerical value reflecting the growth state of the monitoring object, and the larger the numerical value of the state coefficient is, the better the corresponding growth state of the monitoring object is; wherein, alpha 1, alpha 2 and alpha 3 are all proportional coefficients, and alpha 1 > alpha 2 > alpha 3 > 1; the state threshold ZTmin is obtained through the storage module, and the state coefficient ZT is compared with the state threshold ZTmin: if the state coefficient ZT is smaller than the state threshold ZTmin, judging that the growth state of the monitored object is abnormal, sending an abnormal processing signal to a real-time monitoring platform by a state monitoring module, and sending the abnormal processing signal to a mobile phone terminal of a manager after the real-time monitoring platform receives the abnormal processing signal; if the state coefficient ZT is larger than or equal to the state threshold ZTmin, judging that the growth state of the monitored object is normal, sending a data updating signal to the real-time monitoring platform by the state detection module, sending the data updating signal to the data processing module after the data updating signal is received by the real-time monitoring platform, and acquiring new live pig growth parameters for processing and updating a data source of the weight estimation model by the data processing module after the data updating signal is received by the data processing module; and (3) monitoring and analyzing the growth state of the monitored object, comprehensively analyzing a plurality of parameters of the monitored image to obtain a state coefficient, feeding back the growth state of the live pigs according to the numerical value of the state coefficient, updating a model data source when the growth state is normal, and carrying out exception handling when the growth state is abnormal.
Example two
As shown in fig. 2, the live pig breeding weight real-time monitoring method based on image recognition comprises the following steps:
step one: and (3) collecting point cloud images of live pigs: the method comprises the steps of selecting a collected live pig and a collection environment, and scanning the live pig in the collection environment through AzureKinect to obtain three-dimensional point cloud data of the live pig;
step two: processing three-dimensional point cloud data of live pigs to obtain growth parameters i, i=1, 2, …, n and n of the live pigs, selecting a plurality of live pigs as specimen objects, collecting and processing weight values of the specimen objects and numerical values of the growth parameters i, and marking the growth parameters as irrelevant parameters or related parameters;
step three: body weight monitoring analysis was performed on live pig body weight: marking a live pig subjected to weight monitoring analysis as a monitoring object, acquiring and processing a point cloud image of the monitoring object to obtain a weight estimated value TZi and an estimated deviation coefficient corresponding to the related parameter i, and judging whether an estimated analysis result meets the requirement or not through the estimated deviation coefficient;
step four: when the estimated analysis result of the monitoring object does not meet the requirement, the growth state of the monitoring object is monitored and analyzed to obtain a state coefficient ZT of the monitoring object, and the reason that the estimated result of the weight does not meet the requirement is judged according to the value of the state coefficient ZT.
Live pig breeding weight real-time monitoring system based on image recognition gathers live pig's some cloud image at the during operation: the method comprises the steps of selecting a collected live pig and a collection environment, and scanning the live pig in the collection environment through AzureKinect to obtain three-dimensional point cloud data of the live pig; processing three-dimensional point cloud data of live pigs to obtain growth parameters i, i=1, 2, …, n and n of the live pigs, selecting a plurality of live pigs as specimen objects, collecting and processing weight values of the specimen objects and numerical values of the growth parameters i, and marking the growth parameters as irrelevant parameters or related parameters; marking a live pig subjected to weight monitoring analysis as a monitoring object, acquiring and processing a point cloud image of the monitoring object to obtain a weight estimated value TZi and an estimated deviation coefficient corresponding to the related parameter i, and judging whether an estimated analysis result meets the requirement or not through the estimated deviation coefficient; when the estimated analysis result of the monitoring object does not meet the requirement, the growth state of the monitoring object is monitored and analyzed to obtain a state coefficient ZT of the monitoring object, and the reason that the estimated result of the weight does not meet the requirement is judged according to the value of the state coefficient ZT.
The foregoing is merely illustrative of the structures of this invention and various modifications, additions and substitutions for those skilled in the art can be made to the described embodiments without departing from the scope of the invention or from the scope of the invention as defined in the accompanying claims.
The formulas are all formulas obtained by collecting a large amount of data for software simulation and selecting a formula close to a true value, and coefficients in the formulas are set by a person skilled in the art according to actual conditions; such as: formula zt=α1sz/(α2sbs+α3xs); collecting a plurality of groups of sample data by a person skilled in the art and setting a corresponding state coefficient for each group of sample data; substituting the set state coefficient and the acquired sample data into a formula, forming a ternary one-time equation set by any three formulas, screening the calculated coefficient, and taking an average value to obtain values of alpha 1, alpha 2 and alpha 3 which are 3.74, 2.97 and 2.65 respectively;
the size of the coefficient is a specific numerical value obtained by quantizing each parameter, so that the subsequent comparison is convenient, and the size of the coefficient depends on the number of sample data and the corresponding harmful coefficient is preliminarily set for each group of sample data by a person skilled in the art; as long as the proportional relation between the parameter and the quantized value is not affected, for example, the state coefficient is directly proportional to the value of the color data.
In the description of the present specification, the descriptions of the terms "one embodiment," "example," "specific example," and the like, mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiments or examples. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
The preferred embodiments of the invention disclosed above are intended only to assist in the explanation of the invention. The preferred embodiments are not intended to be exhaustive or to limit the invention to the precise form disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best understand and utilize the invention. The invention is limited only by the claims and the full scope and equivalents thereof.
Claims (9)
1. The live pig breeding weight real-time monitoring system based on image recognition is characterized by comprising a real-time monitoring platform, wherein the real-time monitoring platform is in communication connection with an image acquisition module, a data processing module, a model generation module, a state monitoring module and a storage module;
the image acquisition module is used for acquiring point cloud images of live pigs: scanning the live pigs in the selected collected live pigs and the collecting environment through AzureKinect to obtain three-dimensional point cloud data of the live pigs, and sending the three-dimensional point cloud data of the live pigs to a data processing module;
the data processing module is used for processing three-dimensional point cloud data of live pigs to obtain growth parameters i of the live pigs, i=1, 2, …, n and n are positive integers, a plurality of live pigs are selected as specimen objects, the growth parameters i of the specimen objects are subjected to association analysis, the growth parameters i are marked as irrelevant parameters i or association parameters i, and the association parameters i of the specimen objects are sent to the model generating module;
the model generation module is used for generating a weight estimation model through the associated parameters of the specimen object; the weight of the live pigs is monitored and analyzed through the weight estimation model, a weight output value is generated and output when the estimated and analyzed result meets the requirement, and an estimated abnormal signal is sent to the state monitoring module through the real-time monitoring platform when the estimated and analyzed result does not meet the requirement;
the state monitoring module is used for monitoring and analyzing the growth state of the monitored object.
2. The live pig breeding weight real-time monitoring system based on image recognition according to claim 1, wherein the specific process of performing the correlation analysis on the growth parameter i of the specimen object comprises the following steps: the method comprises the steps of collecting a weight value of a specimen object and a numerical value of a growth parameter i, weighing the weight value of the specimen object, setting a weight range of the specimen object, dividing a maximum value and a minimum value of the weight range of the specimen object into a plurality of individual weight sections, marking the numerical value of the growth parameter i of the specimen object as SZix when the weight value of the specimen object is a minimum boundary value of the weight section, marking the numerical value of the growth parameter i of the specimen object as SZid when the weight value of the specimen object is a maximum boundary value of the weight section, marking the absolute value of a difference value between SZid and SZix as a change value BHi of the growth parameter i, acquiring a change threshold value BHmin through a storage module, comparing the change value BHi of the growth parameter i with the change threshold value BHmin, and marking the growth parameter as an irrelevant parameter i or a related parameter i through a comparison result.
3. The live pig breeding weight real-time monitoring system based on image recognition according to claim 2, wherein the specific process of comparing the change value BHi of the growth parameter i with the change threshold BHmin comprises the following steps: if the variation value BHi is larger than or equal to the variation threshold value BHmin, marking the growth parameter i as an influence parameter i corresponding to the weight interval; if the variation value BHi is smaller than the variation threshold value BHmin, marking the growth parameter i as a random parameter i corresponding to the weight interval; marking the number ratio of the times that the growth parameter i is marked as an influence parameter in the weight range to the number of weight intervals as an influence coefficient, acquiring an influence threshold value through a storage module, and comparing the influence coefficient with the influence threshold value: if the influence coefficient is smaller than the influence threshold, marking the corresponding growth parameter i as an irrelevant parameter i of the specimen object; and if the influence coefficient is greater than or equal to the influence threshold, marking the corresponding growth parameter i as the associated parameter i of the specimen object.
4. The live pig breeding weight real-time monitoring system based on image recognition according to claim 3, wherein the specific process of generating the weight estimation model by the model generation module through the relevant parameter i of the specimen object comprises the following steps: summing all the SZix of the sample objects in the weight interval to obtain a low average value DJi of the associated parameter i, summing all the SZid of the sample objects in the weight interval to obtain a high average value GJi of the associated parameter i, marking the difference value between the maximum boundary value and the minimum boundary value of the weight interval as a body difference value TC, obtaining an influence coefficient YXi of the associated parameter i through the high average value GJi, the low average value DJi and the body difference value TCi, and generating a weight estimation model through the influence coefficient YXi of all the associated parameter i.
5. The real-time monitoring system for pig farm weight based on image recognition according to claim 4, wherein the specific process of the weight estimation model for carrying out weight monitoring analysis on pig weight comprises: marking a live pig subjected to weight monitoring analysis as a monitoring object, acquiring and processing a point cloud image of the monitoring object to obtain a numerical value of an associated parameter i of the monitoring object, marking the numerical value as JCi, and obtaining a weight estimated value TZi corresponding to the associated parameter i through a formula TZi= | JCi-DJi | YXi + BJx, wherein BJx is a minimum boundary value corresponding to a weight interval; and establishing a weight set of weight estimation values TZi corresponding to all the associated parameters i, performing variance calculation on the weight set to obtain an estimation deviation coefficient, acquiring an estimation deviation threshold value through a storage module, comparing the estimation deviation coefficient with the estimation deviation threshold value, and judging whether an estimation analysis result meets the requirement or not through a comparison result.
6. The live pig farming weight real-time monitoring system based on image recognition according to claim 5, wherein the specific process of comparing the estimated deviation coefficient with the estimated deviation threshold comprises: if the estimated deviation coefficient is smaller than the estimated deviation threshold, judging that the estimated analysis result meets the requirement, summing the weight estimated values TZi of all the associated parameters i, averaging to obtain a weight output value, and outputting the weight output value to a real-time monitoring platform; if the estimated deviation coefficient is larger than or equal to the estimated deviation threshold, judging that the estimated analysis result does not meet the requirement, sending an estimated abnormal signal to a real-time monitoring platform by the weight estimation model, and sending the estimated abnormal signal to a state monitoring module after the estimated abnormal signal is received by the real-time monitoring platform.
7. The live pig breeding weight real-time monitoring system based on image recognition according to claim 6, wherein the specific process of the state monitoring module for monitoring and analyzing the growth state of the monitored object comprises: the method comprises the steps of performing image shooting on a monitoring object to obtain a monitoring image, amplifying the monitoring image into a pixel grid image, performing gray level conversion to obtain gray level values of the pixel grid, summing the gray level values of all the pixel grids, taking an average value to obtain color data SZ of the monitoring object, and performing body surface analysis on the monitoring object: the method comprises the steps that a gray threshold value is obtained through a storage module, pixel grids with the gray value larger than the gray threshold value are marked as spot grids, a pixel grid area formed by mutually contacted spot grids is marked as a spot area, the number of spot areas and the total number of the pixel grids in the spot area are obtained and respectively marked as BS and XS, and a state coefficient ZT of a monitoring object is obtained through numerical calculation of SZ, BS and XS; the state threshold value ZTmin is obtained through the storage module, the state coefficient ZT is compared with the state threshold value ZTmin, and whether the growth state of the live pigs is normal or not is judged according to a comparison result.
8. The live pig breeding weight real-time monitoring system based on image recognition according to claim 7, wherein the specific process of comparing the state coefficient ZT with the state threshold ZTmin comprises: if the state coefficient ZT is smaller than the state threshold ZTmin, judging that the growth state of the monitored object is abnormal, sending an abnormal processing signal to a real-time monitoring platform by a state monitoring module, and sending the abnormal processing signal to a mobile phone terminal of a manager after the real-time monitoring platform receives the abnormal processing signal; if the state coefficient ZT is larger than or equal to the state threshold ZTmin, the growth state of the monitoring object is judged to be normal, the state detection module sends a data updating signal to the real-time monitoring platform, the real-time monitoring platform receives the data updating signal and then sends the data updating signal to the data processing module, and the data processing module acquires new live pig growth parameters for processing after receiving the data updating signal and updates a data source of the weight estimation model.
9. The working method of the live pig breeding weight real-time monitoring system based on image recognition according to any one of claims 1 to 8, comprising the following steps:
step one: and (3) collecting point cloud images of live pigs: the method comprises the steps of selecting a collected live pig and a collection environment, and scanning the live pig in the collection environment through AzureKinect to obtain three-dimensional point cloud data of the live pig;
step two: processing three-dimensional point cloud data of live pigs to obtain growth parameters i, i=1, 2, …, n and n of the live pigs, selecting a plurality of live pigs as specimen objects, collecting and processing weight values of the specimen objects and numerical values of the growth parameters i, and marking the growth parameters as irrelevant parameters or related parameters;
step three: body weight monitoring analysis was performed on live pig body weight: marking a live pig subjected to weight monitoring analysis as a monitoring object, acquiring and processing a point cloud image of the monitoring object to obtain a weight estimated value TZi and an estimated deviation coefficient corresponding to the related parameter i, and judging whether an estimated analysis result meets the requirement or not through the estimated deviation coefficient;
step four: when the estimated analysis result of the monitoring object does not meet the requirement, the growth state of the monitoring object is monitored and analyzed to obtain a state coefficient ZT of the monitoring object, and the reason that the estimated result of the weight does not meet the requirement is judged according to the value of the state coefficient ZT.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310192317.9A CN116235799A (en) | 2023-03-02 | 2023-03-02 | Live pig breeding weight real-time monitoring system based on image recognition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310192317.9A CN116235799A (en) | 2023-03-02 | 2023-03-02 | Live pig breeding weight real-time monitoring system based on image recognition |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116235799A true CN116235799A (en) | 2023-06-09 |
Family
ID=86623831
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310192317.9A Withdrawn CN116235799A (en) | 2023-03-02 | 2023-03-02 | Live pig breeding weight real-time monitoring system based on image recognition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116235799A (en) |
-
2023
- 2023-03-02 CN CN202310192317.9A patent/CN116235799A/en not_active Withdrawn
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111460990B (en) | Big data-based alpine pastoral area grassland insect pest monitoring and early warning system and method | |
| CN112213962A (en) | Intelligent feeding system and method based on growth model and sonar feedback | |
| CN107610122B (en) | Micro-CT-based single-grain cereal internal insect pest detection method | |
| CN111680587B (en) | Multi-target tracking-based chicken flock activity real-time estimation method and system | |
| CN116034904B (en) | Pig health monitoring system and method based on track type inspection robot | |
| CN115613128A (en) | Intelligent control system for crystal growth | |
| CN116485615A (en) | Method for measuring and calculating threshold concentration of ecological toxicity effect of environmental pollutants | |
| CN110287902B (en) | Livestock and poultry survival detection method, device, equipment and computer program product | |
| CN116627082A (en) | Intelligent management system for live pig breeding suitable for stock quantity checking | |
| CN115661650A (en) | Farm management system based on data monitoring of Internet of things | |
| CN115342859A (en) | Multifunctional grain condition detection system and detection method | |
| CN116453060A (en) | Intelligent pesticide application decision feasibility assessment system for plant planting | |
| CN116235799A (en) | Live pig breeding weight real-time monitoring system based on image recognition | |
| CN116868746B (en) | Intelligent irrigation and fertilization control system based on data analysis | |
| CN117406593A (en) | Intelligent data processing method and system applied to fishery management | |
| CN117172952A (en) | Agricultural disaster monitoring system based on Internet of things and remote sensing technology | |
| CN110298856B (en) | Non-contact type fish body length estimation method | |
| CN115342937B (en) | Temperature anomaly detection method and device | |
| CN115777560A (en) | Intelligent sow feeding system based on machine vision analysis technology | |
| CN115914560A (en) | Intelligent accurate feeding method and device for sows, electronic equipment and storage medium | |
| KR20210069987A (en) | Crop growth diagnosis system and method using image | |
| CN118033071A (en) | Aquaculture water quality monitoring system based on artificial intelligence | |
| WO2022046657A1 (en) | Method and system of analyzing plant image data and projecting plant growth and health status | |
| CN113545296B (en) | Closed livestock shed is with excrement and automatic clear excrement system of scraping in underground | |
| CN207937840U (en) | A kind of agricultural breeding environmental monitoring system based on big data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20230609 |
|
| WW01 | Invention patent application withdrawn after publication |