CN117893507A - Tea condition analysis method, tea condition analyzer and electronic equipment - Google Patents

Abstract

The invention provides a tea condition analysis method, a tea condition analyzer and electronic equipment, and relates to the field of tea condition analysis, comprising the following steps: s1, acquiring images and depth information of a plurality of theaters; s2, acquiring internal information of a plurality of theaters; s3, determining the plane positions and the outlines of a plurality of tea buds; s4, estimating the sizes of a plurality of theaters according to the depth information; s5, determining the space positions of a plurality of theaters according to the positions of the images, the depth information and the internal information acquired by the tea condition analyzer; s6, determining tea conditions of the plurality of tea leaves according to the plane position, the outline, the size, the space position and the internal information. According to the invention, through non-contact analysis of bud conditions and nondestructive detection, damage to tea leaves by a traditional detection method is avoided, the detection accuracy and efficiency are improved, the harvesting time and the tea growth condition can be accurately judged, and workers are guided to accurately pick. Meanwhile, the quality of the tea leaves can be quantitatively graded, so that the quality and market value of the tea leaves can be improved.

Description

Tea condition analysis method, tea condition analyzer and electronic equipment

Technical Field

The invention relates to the field of tea condition analysis, in particular to a tea condition analysis method, a tea condition analyzer and electronic equipment.

Background

In the tea production process, theaflavin is an important production object. The quality judgment of the tea leaves mainly depends on the form and the content of internal substances of the tea leaves, however, the existing picking and grading links depend on manual resolution, and equipment and standards for quantitative qualitative analysis are lacking. This results in the occurrence of the following problems:

1. The quality of the tea leaves cannot be accurately judged on the tea field site, and the quality of the tea leaves is influenced.

2. The lack of analytical means to quantify and quantify the amount of substances in theaters by direct detection rather than to a dedicated test site makes fine pricing difficult.

3. Tea farmers cannot be guided to accurately judge the tea leaf harvesting time, and the tea leaf production plan and input-output maximization are affected.

Therefore, the portable tea analyzer is used for analyzing the condition of the tea leaves, and has important practical significance for solving the problems.

Disclosure of Invention

The invention aims to solve the technical problems that the quality and the quality of tea leaves cannot be accurately judged on a tea field site, and the portable software and hardware and method for analyzing the tea conditions are lacking, so that the tea production plan is influenced and the input and output are maximized.

A tea plot analysis method comprising:

S1, acquiring images and depth information of a plurality of theaters; the image and the depth information are formed by shooting the plurality of theaters through an RGBD camera;

The tea plot analysis method further comprises:

S2, acquiring internal information of the plurality of theaters, wherein the internal information is formed by shooting the plurality of theaters through a hyperspectral camera, and the RGBD camera and the hyperspectral camera are integrated into a tea analyzer through a customized outer box;

s3, inputting the image to a trained tea bud recognition model based on a deep learning algorithm, completing example segmentation, and determining the plane positions and the contours of the tea buds;

s4, estimating the sizes of the plurality of theaters according to the depth information;

s5, determining the space positions of the plurality of theaters according to the positions of the image, the depth information and the internal information acquired by the tea condition analyzer;

s6, determining tea conditions of the plurality of tea leaves according to the plane position, the outline, the size, the space position and the internal information.

Further, the method further comprises the following steps: s701, determining the quantity of the tea-leaf plucking quantity according to the tea conditions of the plurality of tea leaves so as to form picking or classifying opinions according to the quantity of the tea-leaf plucking quantity.

Optionally, S702, classifying the theaters according to the tea conditions to form classification information;

s801, generating a tea production report according to the grading information, the tea production information and the time information.

Further, S802, a theanine transaction report is generated according to the ranking information, the time information and the historical quantized transaction data.

Further, the internal information is chlorophyll grading and distribution information of the theaflavin.

Further, the tea bud recognition model is formed according to YOLOV training.

Optionally, S301, adjusting the technical parameters of the number of categories, the size of the anchor frame, the loss function, the learning rate, the batch size, and the number of training rounds of YOLOV, so as to adapt to the training of the tea bud recognition model.

Further, S302, training set expansion is performed on the images of the plurality of theaters and the depth information based on transfer learning or over-sampling to provide a trained sample of the tea bud recognition model.

The embodiment of the invention provides a tea analyzer, which comprises an outer box, a hyperspectral camera, an RGBD camera and a controller, wherein the hyperspectral camera, the RGBD camera and the controller are accommodated in the outer box, the outer box is made of lightweight materials and is of a portable structure, and the controller is used for realizing the tea analysis method by controlling the hyperspectral camera and the RGBD camera.

Another embodiment of the present invention provides an electronic device, including:

A processor, a memory and a computer program stored on the memory and executable on the processor, which when executed implements the tea profile analysis method described above.

Compared with the prior art, the invention has the following advantages and beneficial effects:

1. According to the tea condition analysis method, the tea condition is analyzed in a non-contact mode, nondestructive detection is achieved, damage to tea leaves by a traditional detection method is avoided, and detection accuracy and efficiency are improved. The portable design can guide manual accurate picking and improve the quality of tea. By analyzing the morphology and the content of the internal substances of the tea, the harvesting time and the growth condition of the tea can be accurately judged, and workers are guided to accurately pick. Meanwhile, the quality of the tea leaves is quantitatively graded, so that the quality and market value of the tea leaves can be improved;

2. According to the tea analyzer, the bud condition is analyzed in a non-contact mode, nondestructive detection is achieved, damage to tea leaves by a traditional detection method is avoided, and detection accuracy and efficiency are improved. The portable design can guide manual accurate picking and improve the quality of tea. By analyzing the morphology and the content of the internal substances of the tea, the harvesting time and the growth condition of the tea can be accurately judged, and workers are guided to accurately pick. Meanwhile, the quality of the tea leaves is quantitatively graded, so that the quality and market value of the tea leaves can be improved;

3. According to the electronic equipment, the bud condition is analyzed in a non-contact mode, nondestructive detection is achieved, damage to tea leaves by a traditional detection method is avoided, and detection accuracy and efficiency are improved. The portable design can guide manual accurate picking and improve the quality of tea. By analyzing the morphology and the content of the internal substances of the tea, the harvesting time and the growth condition of the tea can be accurately judged, and workers are guided to accurately pick. Meanwhile, the quality of the tea leaves can be quantitatively graded, so that the quality and market value of the tea leaves can be improved.

Drawings

The drawings are included to provide a better understanding of the present invention and are not to be construed as limiting the invention. Wherein:

FIG. 1 is a flow diagram of a tea plot analysis method according to one embodiment of the application;

FIG. 2 is a schematic diagram of a tea analyzer software architecture according to one embodiment of the application;

FIG. 3 is a schematic view of a tea condition analyzer according to an embodiment of the present application;

fig. 4 is a schematic diagram of an electronic device according to one embodiment of the application.

Reference numerals:

100-tea analyzer, 101-case, 102-hyperspectral camera, 103-RGBD camera, 104-controller, 200-electronic device, 201-processor, 202-memory.

Detailed Description

Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to like or similar elements or elements having like or similar functions throughout. The embodiments described below by referring to the drawings are illustrative and intended to explain the present invention and should not be construed as limiting the invention.

As shown in fig. 1, the present invention proposes a tea plot analysis method, which can be implemented by a tea plot analyzer 100 or an electronic device 200, comprising:

s1, acquiring images and depth information of a plurality of theaters; wherein the image and the depth information are formed by shooting the plurality of theaters through the RGBD camera 103;

The tea plot analysis method further comprises:

S2, acquiring internal information of the plurality of theaters, wherein the internal information is formed by shooting the plurality of theaters through a hyperspectral camera 102, and the RGBD camera 103 and the hyperspectral camera 102 are integrated into a tea analyzer 100 through a customized outer box 101;

s3, inputting the image to a trained tea bud recognition model based on a deep learning algorithm, completing example segmentation, and determining the plane positions and the contours of the tea buds;

s4, estimating the sizes of the plurality of theaters according to the depth information;

S5, determining the spatial positions of the plurality of theaters according to the positions of the image, the depth information and the internal information acquired by the tea condition analyzer 100;

s6, determining tea conditions of the plurality of tea leaves according to the plane position, the outline, the size, the space position and the internal information.

In some embodiments, the invention resides in: through multi-sensor analysis and portable design, the non-contact analysis of tea conditions is realized, and the efficiency and quality of tea production are improved; through the application of double scenes, the method can be used for field production and quantitative transaction; the production quantitative analysis and the trade constant standard improve the efficiency of tea production and sales.

The tea analyzer 100 is of portable design and is convenient to carry and operate. The inter-integrated hyperspectral camera 102 and RGBD camera 103 (which may also include more sensors) are able to acquire high definition images and depth information of theaflavins. At the same time, the analyzer is also equipped with a high performance processor 201 and memory 202 for processing and storing the acquired data. Therefore, workers or technicians can accurately judge the quality of the tea leaves on the tea field.

The software of the portable tea analyzer 100 is developed based on a deep learning algorithm, and can automatically identify and count tea shoots, and estimate the size of tea shoots and chlorophyll content.

And the manual handheld analyzer photographs the un-picked tea green groups in the tea garden, and simultaneously acquires images, depth and spectrum information. The software can establish a tea bud recognition model through a deep learning algorithm, perform instance segmentation, automatically acquire the position information of tea buds and determine the tea bud condition. The size of the tea leaf is estimated by combining the depth information, and the grading of the inner components (mainly chlorophyll) of the tea leaf is realized by combining a multispectral camera. Finally, counting to obtain the number (unit area) of the tea leaves which can be picked, and accurately picking by workers according to the analysis condition. Therefore, the tea production plan can be more effectively promoted, and the input and output are maximized.

As shown in FIG. 1, after the tea leaf picture is acquired, a tea bud recognition model can be established through a deep learning algorithm to perform instance segmentation. Example segmentation is a high-level task of object detection that can not only identify objects in an image (e.g., tea shoots), but also give accurate boundaries for each object, providing accurate location and contour shape of tea shoots. The tea leaves which are not picked in the tea field are collected in a overlooking mode. For the collected tea leaves, the tea leaves are laid flat by fixing the tea analyzer 100 at a fixed position. And judging whether the tea leaves are standard tea leaves or butt-clamped (poor-quality) tea leaves according to the calculated plane positions and the calculated outlines.

The depth image data set of the tea leaf samples also needs to be constructed according to the estimated sizes of the tea leaves according to the depth information. These data sets may include samples of theanine under different types, different growth environments, different picking times, etc., and the data is cleaned, standardized, and enhanced to achieve labeling. The size extraction of the depth image may be implemented using a BP neural network, a Bayesian classifier, or a depth neural network, with one example being using PointNet or PointNet ++ libraries of PyTorch.

According to the position of the image, the depth information and the internal information acquired by the tea condition analyzer 100, the spatial positions of the plurality of theaters may be determined by calibrating the hyperspectral camera 102 and the RGBD camera 103 respectively and then comprehensively analyzing the spatial positions of the theaters, or may be calculated by calibrating only one of the cameras, for example, by using the RGBD camera 103 to complete by using a Zhang Zhengyou calibration method, which may specifically be: first, a suitable theanine marker, such as a road sign, is selected, with a two-dimensional code, checkerboard, or other pattern of overt features applied thereto. These patterns will be used to determine the intrinsic and distortion parameters of the camera. Then, calibration images are taken, i.e., the theanine markers are placed at different positions and angles around the theanine, and images acquired by the RGBD camera 103 are taken from a plurality of angles. These images will be used to calculate the internal parameters and distortion parameters of the camera. And extracting characteristic points, namely extracting the characteristic points on the theanine markers in the shot calibration images. These feature points will be used to calculate internal parameters and distortion parameters of the camera. Then, the internal parameters of the camera (such as focal length, principal point coordinates, etc.) are calculated according to the photographed calibration image and the feature point information, i.e., using Zhang Zhengyou calibration method or other camera calibration method. And performing correction cutting on the next step, namely performing distortion correction on the acquired tea green image according to the calculated internal parameters so as to eliminate the distortion effect in the image. And finally determining the space positions of the tea leaves, namely determining the space positions of a plurality of tea leaves in the corrected tea leaf image through the technologies of feature extraction, matching and the like. This can be achieved by performing steps such as segmentation, edge detection, morphological processing, etc. on the image.

Further comprises: s701, determining the quantity of the tea-leaf plucking quantity according to the tea conditions of the plurality of tea leaves so as to form picking or classifying opinions according to the quantity of the tea-leaf plucking quantity.

In this embodiment, the tea analyzer 100 of the present invention is used in tea fields, and the number of the tea leaves (unit area) which can be picked up can be automatically counted by software, and then the workers can accurately pick or sort the tea leaves according to the analysis conditions.

Specifically, the position and shape of the theaters can be accurately identified and located through analysis of the image and depth information. The quality and nutritional status of the theaflavins, such as whether the leaves are mature, pest and disease conditions, etc., can be evaluated in combination with the positional data of the internal information. By comprehensively analyzing the tea conditions of a plurality of tea leaves, the quantity of the tea leaves which can be picked can be determined.

Depending on the number of theaters that can be picked, a picking or sorting opinion may be provided. For example, for a pickable tea leaf, the time, manner, and tools of picking, etc. may be suggested; for non-pluckable theaters, classification or no treatment may be recommended, and harvesting after a period of maturation. In addition, the tea leaves can be provided with corresponding classification comments according to the quality and the nutrition condition of the tea leaves, for example, the tea leaves with high quality and the tea leaves with low quality are picked or treated respectively.

By analyzing the tea conditions of a plurality of tea leaves, the number of pickable tea leaves can be determined, and picking or sorting comments can be provided according to the number of pickable tea leaves. The technical scheme is beneficial to improving the production efficiency and quality control level, reducing the labor cost and providing an innovative solution for tea production and management. Meanwhile, an improvement thought can be provided for tracing, quality assurance and safe production of tea.

S702, classifying the tea leaves according to the tea conditions to form classification information;

s801, generating a tea production report according to the grading information, the tea production information and the time information.

Specifically, an analyzer is erected manually or mechanically, the collected tea leaves are selected randomly and laid on a scale, and identification and judgment are carried out on the picked tea leaves. The software carries out classification discrimination and proportion calculation (butt-clamp and normal tea leaf) on tea leaves through the same algorithm model, and simultaneously carries out tea bud counting, size estimation and chlorophyll estimation. Finally, the condition of the batch of tea leaves is qualitatively classified.

In some embodiments, portable tea analyzer 100 uploads the raw data and processing results obtained to a database for use in tea production management and quantitative data for tea leaf transactions. The database may also include a data processing module for processing and analyzing tea conditions to generate tea production reports and tea leaf transaction reports. Meanwhile, the database can also provide a visual interface, so that a user can conveniently inquire and analyze data.

The tea leaves are classified according to the image and depth information acquired by the tea information analyzer 100. Classification can be performed in a variety of ways, such as classification based on the shape, color, size, thickness, etc. characteristics of the tea leaves. The classification model can be trained to automatically classify the tea leaves through techniques such as machine learning, artificial intelligence and the like. The graded theaflavins may form grading information such as the number, quality, etc. of the different grades of theaflavins.

And generating a tea production report by combining the tea production information. Tea production information may include a variety of factors such as tea tree variety, planting environment, fertilization conditions, picking time, etc. By comprehensively analyzing these factors, a tea production report with reference value can be generated. The report may include a variety of content, such as different grades of theabrownish yield, production costs, quality characteristics, and the like. These reports may provide decision support for tea producers, helping them to make more rational production plans and management strategies.

And analyzing and evaluating the tea production report by combining the time information. The time information may include seasonal variations, climate conditions, market demand, and the like. Through analysis and evaluation of these factors, correlations and trends between tea production and market demand can be derived.

Therefore, by comprehensively analyzing the classification information, the tea production information and the time information, a production report with reference value can be obtained, intelligent decision support is provided for tea production and management, and the production efficiency, the cost and the quality control level are improved. Meanwhile, the method can also be extended to play roles in tracing the source of tea, guaranteeing the quality and producing the safety.

S802, generating a tea leaf transaction report according to the grading information, the time information and the historical quantized transaction data.

In some embodiments, a theaflavin transaction report is generated based on the ranking information, the time information, and the historical quantized transaction data. The transaction report may include a variety of information such as current market price, historical transaction trends, supply and demand relationships, and the like. Such information may provide decision support for tea producers, merchants, and consumers to assist them in making more intelligent transaction decisions. In addition, the method can also be combined with tea quality evaluation results to provide an intelligent solution for tea production and management.

Thus, by comprehensively utilizing technologies such as image processing, computer vision, machine learning, artificial intelligence and the like, automatic and intelligent tea processing, grading and transaction report generation are realized. The grade and quality of the tea leaf can be accurately determined through analysis of detailed information such as appearance characteristics, shape and surface texture of the tea leaf. Meanwhile, by combining time information and historical quantized transaction data, a transaction report with reference value can be generated, and decision support is provided for tea producers, merchants and consumers. The technical scheme is also beneficial to improving the production efficiency, reducing the cost and improving the quality control level.

The internal information is chlorophyll grading and distribution information of the tea leaves.

Further, specifically, the internal information is formed by photographing the plurality of theaters by the hyperspectral camera 102, specifically, the following manner may be implemented: the hyperspectral image is preprocessed, including noise filtering, image normalization and the like, so that contrast and definition of the image are enhanced. And then carrying out deep analysis on the preprocessed image by using a PCA algorithm. The PCA algorithm effectively extracts the principal features in the image and projects the data into a new coordinate system consisting of the principal feature vectors of the data. After the PCA analysis is completed, a new set of principal component images is obtained. These principal component images not only contain most of the information in the original image, but are independent of each other. Next, the principal component of the theaflavin having the highest correlation with chlorophyll content may be selected and further processed using regression analysis. Regression analysis can provide us with a quantitative relationship between chlorophyll content and the selected principal component. And finally, calculating the chlorophyll content value of each point of the tea leaves according to the regression analysis result. The calculations are integrated together to form a detailed chlorophyll profile. The distribution map not only can accurately show the chlorophyll content of the local part of the leaf, but also can comprehensively reveal the distribution condition of the chlorophyll of the whole tea leaf.

In particular, chlorophyll is one of important photosynthetic pigments in plants, and is closely related to the nutritional status and growth conditions of plants. Through analysis of chlorophyll distribution conditions, the growth condition, the nutrition level, the quality characteristics and the like of the tea green can be evaluated. According to chlorophyll grading and distribution information, decision support can be provided for tea production and management. For example, for a tea leaf with a lower chlorophyll content, corresponding management measures such as increasing fertilization, improving irrigation, etc. can be taken to promote the growth and development of the tea leaf. For the theaflavin with higher chlorophyll content, the theaflavin can be used as a candidate object of high-quality tea to carry out fine processing and high-end market development.

In some embodiments, the tea bud recognition model is formed according to YOLOV training.

Specifically, using the YOLOv S model in yolov5v7.0 instance partition, CPU speed test is performed at ONNXFP and exported to TensorRTFP for GPU speed test, image size iro=0.01 is 640 and weight_decay=5e-5 has all default settings.

S301, adjusting the technical parameters of the category number YOLOV, the anchor frame size, the loss function, the learning rate, the batch size and the training round number to adapt to training of the tea bud recognition model.

Specifically, in some embodiments, some parameters are set by considering technical parameters of category number, anchor frame size, loss function, learning rate, batch size, and training wheel number.

Category number (Number of classes): the number of categories refers to the number of target categories to be classified in the target detection task. If there are multiple categories of shoots in the dataset, for example, shoots of different varieties or different growth stages, it may be necessary to increase the number of categories to identify these differences. In the tea bud recognition task, the category number can be determined according to actual demands, for example, tea buds, other weeds and the like can be classified into two categories, and the tea buds which are relatively close can be classified.

Anchor frame size (Anchor sizes): the anchor frame size refers to a preset frame used to locate a target frame in target detection. In tea bud identification, a proper anchor frame size is usually selected according to different scenes and task requirements. During network training, a prediction frame is output on the basis of an initial anchor frame, then the prediction frame is compared with a real frame, the difference between the prediction frame and the real frame is calculated, and then network parameters are reversely updated and iterated. Thus, the placement of the initial anchor frame is also a relatively important part. And adopting self-adaption to calculate the optimal anchor frame values in different training sets.

Loss function (Loss function): the loss function is used for calculating the error between the model prediction result and the actual label and guiding the model optimization direction. The loss function is used to optimize the predicted outcome of the model. In the field of object detection, common loss functions include cross entropy loss, ioU loss, gloU loss, and the like. For different tasks and data sets, it may be necessary to adjust the loss function to better optimize the outcome of tea bud recognition.

Learning rate (LEARNING RATE): the learning rate is the step size used to update the model parameters. In tea bud recognition, a proper learning rate is selected according to factors such as task complexity and model performance, and a smaller learning rate is generally used to avoid over-fitting.

Batch size (Batch size): the batch size refers to the number of samples used in each training iteration. In tea bud identification, the choice of batch size can affect model training speed and convergence effect. The batch size is selected in the range of 32 to 256.

Training round number (Epochs): the training round number refers to the total number of model training times. The more training rounds, the more features and rules the model can learn, but at the same time it is also prone to over-fitting. Typically, the number of training rounds is selected in the range of 30 to 100.

The following is an example application:

# import required library

import torch

import torch.nn as nn

import torch.optim as optim

# Definition model

class YourModel(nn.Module)：

def__-int_(selff)：

super(YourModel，self).__init__()

Model architecture definition

def forward(self，x)：

Forward propagation definition #

return x

# Definition super parameter

Learning_rate=0.001# learning rate

Batch_size=64# batch size

Epochs = 50# training wheel number

# Instantiation model, loss function and optimizer

model＝YourModel()

Criterion=nn. Cross EntropyLoss () # loss function selection, task specific

Optimizer = optimizers }, lr = learning_rate, weight_decay =5e-5) # optimizer choice, depending on the particular task

Training model #

for epoch in range(epochs)：

For i, (inputs, labes) inenumerate (train_loader): # data loader, dependent on specific data set

Optimizer zero_grad () # empties the gradient cache

Output = model (inputs) # forward propagation computation output result

Loss = criterion (outputs, labes) # calculate the loss function value

Back-propagation calculated gradient values

The model parameters are updated by the optimizer step () # according to the gradient values

Reference may be made to links https: the secondary development is completed by// gitsub.com/ultralytics/yolov 5/releases, and will not be described here again.

In some embodiments, further comprising: s302, performing training set expansion on the images and the depth information of the plurality of theaters based on transfer learning or over-sampling to provide a training sample of the tea bud recognition model.

In an exemplary technical scheme, a transfer learning technology is adopted, fine adjustment is performed by using a model trained on a data set, and sample data expansion is realized by using a labeled tea tree data set model through transfer learning of scale invariance parameters. Or the small sample data is processed by adopting an oversampling technology, for example, more than 10 tea bud pictures are shot by one machine position to carry out oversampling so as to increase the data volume and improve the performance of YOLOV algorithm.

The invention also provides a tea analyzer 100, which comprises an outer box 101, a hyperspectral camera 102, an RGBD camera 103 and a controller 104, wherein the hyperspectral camera 102, the RGBD camera 103 and the controller 104 are accommodated in the outer box 101, the outer box 101 is made of lightweight materials into a portable structure, and the controller 104 is used for realizing the tea analysis method by controlling the hyperspectral camera 102 and the RGBD camera 103.

The tea analyzer 100 comprises a hyperspectral camera 102, an RGBD camera 103 and other sensors, can realize non-contact analysis of tea conditions, guide manual accurate picking, and improve tea quality.

The outer case 101 is made of a lightweight material to be a portable structure, the lightweight material can be carbon fiber composite material, aluminum alloy, polycarbonate or glass fiber reinforced plastic, the portable structure is shown in fig. 3, the volume and weight of a product are reduced through compact shape and size, and the thin wall design is adopted, the wall thickness of the outer case 101 is 0.1mm-0.3mm, so that workers or technicians can conveniently carry the tea analyzer 100 of the invention to realize the method, and the quality of tea green can be accurately judged on a tea field site, so that the tea production plan and input output are maximized.

In some embodiments, the description may refer specifically to the above examples, which are not repeated here.

Fig. 2 shows a block diagram of an exemplary apparatus suitable for implementing embodiments of the present invention, and the tea analyzer 100 shown is merely an example and should not be construed as limiting the functionality and scope of use of embodiments of the present invention.

The system described above is in the form of a general purpose computing device. The components of the system may also include, but are not limited to: one or more processors 201 or processing units, a system memory 202, and a bus that connects the various system components (including the system memory 202 and processing units), collectively referred to as the controller 104.

The various embodiments of the tea analyzer and techniques described above in this disclosure may be implemented in digital electronic circuitry, integrated circuit systems, fpgas (Field Programmable GATE ARRAY, field programmable gate arrays), ASlC (application-SPECIFIC INTEGRATED circuits ), ASSP (application SPECIFIC STANDARD products, application-specific standard products), SOC (system On chip systems), cplds (Complex Programmable Logic Device, complex programmable logic devices), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor 201, which may be a special purpose or general purpose programmable processor 201, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.

To provide for interaction with a user, the tea condition analyzer and techniques described herein may be implemented on a remote computer having: a display device (e.g., CRT (Cathode-Ray Tube) or LCD (Liquid CRYSTAL DISPLAY) monitor) for displaying information to a user; and a keyboard and pointing device (e.g., a mouse or trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.

The present invention also proposes an electronic device 200 comprising:

the tea analysis method comprises a processor 201, a memory 202 and a computer program stored on the memory 202 and capable of running on the processor 201, wherein the processor 201 realizes the tea analysis method when executing the program.

It should be noted that, as shown in fig. 4, the electronic device 200 in this example is represented in the form of a general-purpose computing device. The components of the electronic device 200 may include, but are not limited to: one or more processors 201 or processing units, a system memory 202, and a bus connecting different system components (including the system memory 202 and processing units).

The electronic device 200 includes a computing unit that can perform various appropriate actions and processes according to a computer program stored in a ROM (read-only memory 202) or a computer program loaded from a storage unit into a RAM (Random Access Memory, random access memory 202). In the RAM, various programs and data required for the operation of the device may also be stored. The computing unit, ROM and RAM are connected to each other by a bus. An I/O (lnput/Output) interface is also connected to the bus.

Various components in the electronic device 200 are connected to the I/O interface, including: an input unit such as a keyboard, a mouse, etc.; an output unit such as various types of displays, speakers, and the like; a storage unit such as a magnetic disk, an optical disk, or the like; and communication units such as network cards, modems, wireless communication transceivers, and the like. The communication unit allows the device to exchange information/data with other devices via a computer network, such as the internet, and/or various telecommunication networks.

The exemplary electronic device 200 used to implement embodiments of the present invention is intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. Electronic device 200 may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smart phones, wearable devices, and other similar computing devices. The components shown in this disclosure, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the inventions described and/or claimed in this disclosure.

The processor 201 is a computing unit that may be a variety of general and/or special purpose processing components with processing and computing capabilities. Some examples of processor 201 include, but are not limited to, a CPU (entral Processing Unit, central processing unit), GPU (Graphic Processing Units, graphics processing unit), various specialized Al (ARTIFICIAL INTELLIGENCE ) computing chips, various computing units running machine learning model algorithms, DSPs (DIGITAL SIGNAL processor ), and any suitable processor 201, controller 104, microcontroller 104, etc. The computing unit performs the various methods and processes described above, such as an imaging method based on a focal plane fit. For example, in some embodiments, the focal plane fitting-based imaging method may be implemented as a computer software program tangibly embodied on a machine-readable medium, such as a storage unit. In some embodiments, part or all of the computer program may be loaded and/or installed onto the device via the ROM and/or the communication unit. One or more steps of the methods described above may be performed when the computer program is loaded into RAM and executed by a computing unit. Alternatively, in other embodiments, processor 201 may be configured to perform the aforementioned methods by any other suitable means (e.g., by means of firmware).

In the context of the present invention, memory 202 may be a machine-readable medium, which may be a tangible medium, that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may also be a machine-readable signal medium or a machine-readable storage medium. The machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, RAM, ROM, EPROM (ELECTRICALLY PROGRAMMABLE READ-only-memory, erasable programmable read-only memory 202) or flash memory 202, an optical fiber, a CD-ROM (Compact Disc Read-only memory 202), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

Program code for carrying out methods of the present invention may be written in any combination of one or more programming languages. These program code may be provided to the processor 201 or controller 104 of a general purpose computer, special purpose computer, or other programmable data processing apparatus such that the program code, when executed by the processor 201 or controller 104, causes the functions/operations specified in the flowchart and/or block diagram to be implemented. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.

Compared with the prior art, the invention has the following advantages and beneficial effects:

1. According to the tea condition analysis method provided by the invention, the tea leaves are prevented from being damaged by the traditional detection method through non-contact bud condition analysis and nondestructive detection, and the detection accuracy and efficiency are improved. The portable design can guide manual accurate picking and improve the quality of tea. By analyzing the morphology and the content of the internal substances of the tea, the harvesting time and the growth condition of the tea can be accurately judged, and workers are guided to accurately pick. Meanwhile, the quality of the tea leaves is quantitatively graded, so that the quality and market value of the tea leaves can be improved;

2. According to the tea condition analyzer, the tea condition is analyzed in a non-contact mode, nondestructive detection is achieved, damage to tea leaves by a traditional detection method is avoided, and detection accuracy and efficiency are improved. The portable design can guide manual accurate picking and improve the quality of tea. By analyzing the morphology and the content of the internal substances of the tea, the harvesting time and the growth condition of the tea can be accurately judged, and workers are guided to accurately pick. Meanwhile, the quality of the tea leaves is quantitatively graded, so that the quality and market value of the tea leaves can be improved;

3. According to the electronic equipment provided by the invention, the damage to tea leaves by a traditional detection method is avoided through non-contact analysis of bud conditions and nondestructive detection, and the detection accuracy and efficiency are improved. The portable design can guide manual accurate picking and improve the quality of tea. By analyzing the morphology and the content of the internal substances of the tea, the harvesting time and the growth condition of the tea can be accurately judged, and workers are guided to accurately pick. Meanwhile, the quality of the tea leaves can be quantitatively graded, so that the quality and market value of the tea leaves can be improved.

In the several embodiments provided in the present invention, it should be understood that the disclosed technology may be implemented in other manners. The above-described embodiments of the apparatus are merely exemplary, and the division of the units, for example, may be a logic function division, and may be implemented in another manner, for example, a plurality of units or components may be combined or may be integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some interfaces, units or modules, or may be in electrical or other forms.

In the description of the present specification, reference to the terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means 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 invention. In this specification, schematic representations of the above terms are not necessarily directed to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, the different embodiments or examples described in this specification and the features of the different embodiments or examples may be combined and combined by those skilled in the art without contradiction.

Furthermore, the terms "first," "second," and the like, are used for descriptive purposes only and are not to be construed as indicating or implying a relative importance or implying a number of technical features being indicated. Thus, a feature defining "a first", "a second" may include at least one such feature, either explicitly or implicitly. In the description of the present invention, the meaning of "plurality" means at least two, for example, two, three, etc., unless specifically defined otherwise.

Claims (10)

1. A tea plot analysis method comprising: s1, acquiring images and depth information of a plurality of theaters; wherein the image and the depth information are formed by shooting the plurality of theaters through an RGBD camera (103); the tea plot analysis method further comprises:

S2, acquiring internal information of the plurality of theaters, wherein the internal information is formed by shooting the plurality of theaters through a hyperspectral camera (102), and the RGBD camera (103) and the hyperspectral camera (102) are integrated into a tea analyzer (100) through a customized outer box (101);

s3, inputting the image to a trained tea bud recognition model based on a deep learning algorithm, completing example segmentation, and determining the plane positions and the contours of the tea buds;

s4, estimating the sizes of the plurality of theaters according to the depth information;

S5, determining the spatial positions of the plurality of theaters according to the positions of the image, the depth information and the internal information acquired by the tea condition analyzer (100);

s6, determining tea conditions of the plurality of tea leaves according to the plane position, the outline, the size, the space position and the internal information.

2. A tea plot analysis method according to claim 1, further comprising: s701, determining the quantity of the tea-leaf plucking quantity according to the tea conditions of the plurality of tea leaves so as to form picking or classifying opinions according to the quantity of the tea-leaf plucking quantity.

3. A tea information analysis method according to claim 1, wherein S702 classifies the tea leaves according to the tea information to form classification information;

s801, generating a tea production report according to the grading information, the tea production information and the time information.

4. A tea plot analysis method according to claim 3, wherein S802 a tea green transaction report is generated based on the ranking information, the time information and historical quantitative transaction data.

5. A tea analysis method according to claim 1, wherein the internal information is chlorophyll classification and distribution information of the theaflavins.

6. A tea plot analysis method as claimed in claim 1, wherein the tea bud recognition model is formed in accordance with YO LOV5 training.

7. A tea analysis method according to claim 6, wherein S301 adjusts YOLOV technical parameters including category number, anchor frame size, loss function, learning rate, batch size, training round number, to adapt to training of the tea bud recognition model.

8. A tea plot analysis method according to claim 1, wherein S302, training set expansion is performed on the images of the plurality of theaters and the depth information based on transfer learning or over-sampling to provide a trained sample of the tea bud recognition model.

9. A tea analyzer (100), comprising an outer case (101) and a hyperspectral camera (102), an RGBD camera (103) and a controller (104) accommodated therein, wherein the outer case (101) is made of lightweight material into a portable structure, and the controller (104) is configured to implement the method as claimed in any one of claims 1 to 8 by controlling the hyperspectral camera (102) and the RGBD camera (103).

10. An electronic device (200), characterized by comprising:

a processor (201), a memory (202) and a computer program stored on the memory (202) and executable on the processor (201), the processor (201) implementing the method according to any one of claims 1-8 when executing the program.