CN116612191B - Automatic positioning method and device for vehicle-mounted picking robot - Google Patents
Automatic positioning method and device for vehicle-mounted picking robot Download PDFInfo
- Publication number
- CN116612191B CN116612191B CN202310875948.0A CN202310875948A CN116612191B CN 116612191 B CN116612191 B CN 116612191B CN 202310875948 A CN202310875948 A CN 202310875948A CN 116612191 B CN116612191 B CN 116612191B
- Authority
- CN
- China
- Prior art keywords
- ganoderma lucidum
- image
- information
- maturity
- gray level
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 51
- 240000008397 Ganoderma lucidum Species 0.000 claims abstract description 240
- 235000001637 Ganoderma lucidum Nutrition 0.000 claims abstract description 240
- 241000222336 Ganoderma Species 0.000 claims abstract description 28
- 230000007613 environmental effect Effects 0.000 claims abstract description 13
- 238000003709 image segmentation Methods 0.000 claims abstract description 8
- 239000013598 vector Substances 0.000 claims description 20
- 230000008569 process Effects 0.000 claims description 13
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 claims description 11
- 239000011159 matrix material Substances 0.000 claims description 10
- 230000000903 blocking effect Effects 0.000 claims description 5
- 230000009466 transformation Effects 0.000 claims description 5
- 238000012545 processing Methods 0.000 abstract description 5
- 241000233866 Fungi Species 0.000 description 16
- 230000011218 segmentation Effects 0.000 description 7
- 238000012549 training Methods 0.000 description 7
- 238000004891 communication Methods 0.000 description 6
- 239000000463 material Substances 0.000 description 6
- 238000004364 calculation method Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 238000003062 neural network model Methods 0.000 description 4
- 239000000843 powder Substances 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 239000000428 dust Substances 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000002093 peripheral effect Effects 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 238000012546 transfer Methods 0.000 description 3
- 238000007405 data analysis Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000004927 fusion Effects 0.000 description 2
- 239000012535 impurity Substances 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 239000002689 soil Substances 0.000 description 2
- 241000894006 Bacteria Species 0.000 description 1
- 229920002101 Chitin Polymers 0.000 description 1
- 241000222341 Polyporaceae Species 0.000 description 1
- 238000010521 absorption reaction Methods 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 229920002678 cellulose Polymers 0.000 description 1
- 239000001913 cellulose Substances 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000012258 culturing Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 210000004602 germ cell Anatomy 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 210000000936 intestine Anatomy 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 210000002784 stomach Anatomy 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/40—Analysis of texture
- G06T7/41—Analysis of texture based on statistical description of texture
- G06T7/45—Analysis of texture based on statistical description of texture using co-occurrence matrix computation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/761—Proximity, similarity or dissimilarity measures
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Computing Systems (AREA)
- Health & Medical Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Mathematical Physics (AREA)
- Artificial Intelligence (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Multimedia (AREA)
- Image Processing (AREA)
Abstract
The invention relates to the technical field of image processing, in particular to an automatic positioning method and device of a vehicle-mounted picking robot, which are implemented by acquiring gray level images, ambient temperature information and humidity information of a ganoderma lucidum area to be detected; image segmentation is carried out on the gray level image to obtain a single ganoderma lucidum image; obtaining average gray level and texture similarity of a single ganoderma lucidum image; calculating the information gain corresponding to any one of the environmental temperature information, the humidity information, the average gray level and the texture similarity of all the single ganoderma lucidum images, and obtaining the maturity according to each information gain and the corresponding characteristics; and when the maturity is greater than the standard maturity, collecting ganoderma spores when the ganoderma in the ganoderma area to be detected is mature ganoderma. The method can accurately evaluate the maturity of the ganoderma lucidum when collecting ganoderma lucidum spores, thereby improving the quality of the collected ganoderma lucidum spores.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to an automatic positioning method and device for a vehicle-mounted picking robot.
Background
Ganoderma is fruiting body of Ganoderma of Polyporaceae. Wherein Ganoderma spore is the seed of Ganoderma lucidum which is the tiny oval germ cell ejected from Ganoderma lucidum pleat in the mature period of Ganoderma lucidum growth. Each ganoderma lucidum spore is only 4-6 microns, is a living organism, has a double-wall structure, is surrounded by hard chitin cellulose, is difficult to fully absorb by human body, and is more suitable for direct absorption by human intestines and stomach after wall breaking.
Because the ganoderma lucidum spores are extremely fine, the ganoderma lucidum spores can fly around along with the air flow, and the ganoderma lucidum spore powder with pure quality is difficult to collect without adopting a special collecting mode.
In the collection of ganoderma lucidum spores, the following points are noted:
1) The yield and quality of Ganoderma spore powder are related to the maturity of Ganoderma;
2) Collecting a proper amount;
3) The high-quality ganoderma lucidum spores are derived from high-quality ganoderma lucidum, so that the high-quality ganoderma lucidum spores are selected for collection before collection.
At present, the method for collecting the ganoderma lucidum spores comprises a mulching film collecting method, a sleeve collecting method and a blower collecting method. The air blower collecting method is that an air blower collecting device is arranged above the planted ganoderma lucidum, and the air blower collecting device is controlled manually to collect ganoderma lucidum spores; it should be noted that, compared with the mulching film collecting method and the sleeve collecting method, the air blower collecting method saves a lot of labor cost and time cost, but the air blower collecting method needs to evaluate the maturity of the ganoderma lucidum manually and collect the ganoderma lucidum, the yield and quality of ganoderma lucidum spore powder are related to the maturity of the ganoderma lucidum, the air blower collecting method may have inaccurate problems due to the fact that the manual evaluation depends on experience of people, meanwhile, dust and impurities are inevitably mixed in ganoderma lucidum spores collected by the air blower collecting method, even soil is sucked into a machine, and the quality of ganoderma lucidum spores is affected by the problems.
Disclosure of Invention
The invention aims to provide an automatic positioning method and device for a vehicle-mounted picking robot, which are used for solving the problem that the quality of ganoderma lucidum spores is low when the ganoderma lucidum spores are collected by the existing method.
In order to solve the technical problems, the invention provides an automatic positioning method of a vehicle-mounted picking robot, which comprises the following steps:
acquiring gray level images, ambient temperature information and humidity information of a ganoderma lucidum area to be detected; image segmentation is carried out on the gray level image to obtain a single ganoderma lucidum image;
obtaining the average gray level of all pixel points in the single ganoderma lucidum image and the texture similarity of the single ganoderma lucidum image;
constructing feature vectors based on the environmental temperature information, the humidity information, the average gray level and the texture similarity of the single ganoderma lucidum image to obtain feature vector groups of all the single ganoderma lucidum images;
according to the feature vector group, calculating information gain corresponding to any feature of the environmental temperature information, the humidity information, the average gray level and the texture similarity in the feature vector group, and calculating to obtain maturity according to each information gain and the corresponding feature;
comparing the maturity with a standard maturity obtained in advance, and collecting ganoderma spores when the maturity is larger than the standard maturity, wherein ganoderma in a ganoderma area to be detected is mature ganoderma; wherein, the pre-acquired standard maturity is the product of the ratio of the weight of the ganoderma lucidum spores of the historical single mature ganoderma lucidum to the average mature weight and the maturity of the acquired standard image.
Preferably, when the image segmentation is performed, the method further comprises the step of judging whether the ganoderma lucidum images are blocked, and if the blocking exists between the ganoderma lucidum images, taking the non-blocked area in the blocked ganoderma lucidum images as a single ganoderma lucidum image.
Preferably, the texture similarity of the single ganoderma lucidum image comprises:
carrying out affine transformation on the single ganoderma lucidum image to obtain an affine image;
acquiring a maximum overlapping area between the affine image and the standard image;
and calculating the texture similarity of the single ganoderma lucidum image and the standard image in the maximum overlapping area.
Preferably, the process of obtaining the texture similarity of the single ganoderma lucidum image comprises the following steps:
acquiring gray level co-occurrence matrixes of all single ganoderma lucidum images and standard images;
randomly selecting at least three characteristic quantities in gray level co-occurrence matrixes corresponding to any single ganoderma lucidum image and standard image, and constructing texture characteristics of any single ganoderma lucidum image and standard image;
and calculating the texture similarity of the texture characteristics of any single ganoderma lucidum image and the standard image.
Preferably, the feature quantity includes at least three of angular second moment, entropy, contrast, and contrast score matrix.
Preferably, the information gain is the difference between the information entropy and the conditional entropy of any feature of the calculated ambient temperature information, humidity information, average gray scale and texture similarity of all the individual ganoderma lucidum images.
Preferably, according to the gain of each information and the corresponding characteristics, the obtaining process of the maturity obtained by calculation comprises the following steps:
and carrying out weighted summation on the information gains and corresponding characteristics in the environmental temperature information, the humidity information, the average gray level and the texture similarity in the single ganoderma lucidum image to obtain the maturity of the single ganoderma lucidum.
Preferably, according to the gain of each information and the corresponding characteristics, the obtaining process of the maturity obtained by calculation comprises the following steps:
calculating the ambient temperature information mean value, the humidity information mean value, the average gray level mean value and the texture similarity mean value of all the single ganoderma lucidum images;
and carrying out weighted summation on the information gains and corresponding characteristics in the ambient temperature information mean value, the humidity information mean value, the average gray level mean value and the texture similarity mean value to obtain the maturity of the ganoderma lucidum region to be detected.
The invention also provides an automatic positioning device of the vehicle-mounted picking robot, which comprises:
a processor; and
the storage is used for storing computer instructions for automatically positioning the vehicle-mounted picking robot, and when the computer instructions are run by the processor, the equipment is enabled to execute the technical scheme of the vehicle-mounted picking robot automatic positioning method.
The invention also provides a computer readable storage medium, on which computer readable instructions for automatic positioning of the vehicle-mounted picking robot are stored, which when executed by one or more processors, implement the technical scheme of the automatic positioning method of the vehicle-mounted picking robot.
The beneficial effects of the invention are as follows:
the invention starts from the image information of the ganoderma lucidum area to be detected and the environmental information, evaluates the maturity condition of the ganoderma lucidum, namely comprehensively characterizes the maturity degree of the ganoderma lucidum from a plurality of dimensions, improves the accuracy of detecting the maturity degree of the ganoderma lucidum, and simultaneously combines with the ratio determined by weight, namely introduces the dimension of the weight of ganoderma lucidum spores, thereby further ensuring the accuracy and reliability of the detection result and realizing the accurate identification of the mature ganoderma lucidum and the collection of ganoderma lucidum spores.
Meanwhile, the importance of the weight occupied by the characteristics is determined by calculating the image information and the information gain of the environment information of the ganoderma lucidum region to be detected, and the information fusion of the image information and the environment information of the ganoderma lucidum region to be detected is carried out, so that the calculated amount of the maturity of the ganoderma lucidum is reduced, the information fusion can be carried out according to the importance of the characteristics, the more important information is highlighted as much as possible, and the maturity of the ganoderma lucidum is estimated more accurately.
Further, the maturity of the single ganoderma lucidum image is compared with the standard maturity, all single ganoderma lucidum images which are not smaller than the standard maturity are reserved, and finally, the ganoderma lucidum spores are collected in a positioning mode aiming at the reserved single ganoderma lucidum images, so that the maturity of ganoderma lucidum can be evaluated more accurately.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of 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 other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
Fig. 1 is a block diagram of a vehicle-mounted picking robot;
FIG. 2 is a schematic view of a simple structure of a vehicle-mounted picking robot;
FIG. 3 is a flow chart of an automatic positioning method of a vehicle-mounted picking robot of the present invention;
FIG. 4 is a flow chart of the texture similarity acquisition process of the present invention;
fig. 5 is a structural view of an automatic positioning device of a vehicle-mounted picking robot of the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the present invention to achieve the preset purpose, the following detailed description is given below of the specific implementation, structure, features and effects of the technical solution according to the present invention with reference to the accompanying drawings and preferred embodiments. In the following description, a different one or another embodiment is not necessarily the same embodiment. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following describes an automatic positioning method of a vehicle-mounted picking robot by taking ganoderma lucidum as an example:
at present, when ganoderma lucidum spores are collected, a blower collection method is generally adopted, a blower collection device is arranged above planted ganoderma lucidum, because the yield and quality of ganoderma lucidum spore powder are related to the maturity of ganoderma lucidum, the maturity is required to be manually evaluated and collected, the inaccuracy problem exists, dust and impurities are inevitably mixed in collected ganoderma lucidum spores, and even soil is sucked into a machine, so that the quality of ganoderma lucidum spores is affected.
Based on the problems, the vehicle-mounted picking robot determines the maturity of the ganoderma lucidum through data analysis of the image data and the environment data information of the ganoderma lucidum, and compares the maturity with the standard maturity of the standard image, so that whether the ganoderma lucidum is mature is determined, automatic positioning collection of ganoderma lucidum spores is realized, and collection of high-quality ganoderma lucidum spores is realized.
The vehicle-mounted picking robot in the embodiment comprises a control system and an execution system. As shown in fig. 1, the control system comprises an image acquisition system, a main controller, a motor driver, a PID regulator and a sensor system; the execution system comprises a suction device and a storage box.
As an embodiment, suction device wherein, including inhaling the material pump, inhale material pipe and expanding mask, wherein inhale material pipe is installed in the embedding of inhaling material pump one side, inhale the one end that the material pipe deviates from inhaling the material pump and install expanding mask for after image acquisition system discerns ripe glossy ganoderma, main control unit sends control command to PID regulator, and PID regulator carries out the collection of glossy ganoderma spore through motor driver control suction device.
Of course, as other embodiments, the suction device of the invention can be similar to the suction device of a dust collector, after the image acquisition system identifies mature ganoderma lucidum, the ganoderma lucidum spores in the identified area are sucked through the suction device, so that the ganoderma lucidum spores are collected and stored in the storage box.
As shown in fig. 1, the image acquisition system and the sensor system are respectively connected with the main controller and used for transmitting the acquired image information and data information to the main controller.
It should be noted that, because the field of view of the image acquisition device is limited, the ganoderma lucidum image can only be acquired in a local area, so that the ganoderma lucidum can be identified and positioned every time the system acquires one image.
The image acquisition system can be arranged on a rod together with the suction device, namely synchronous rotation can be realized through the driving of the motor driver. Of course, as other embodiments, the image acquisition system of the present invention may also be a binocular camera with a pole or a stand on the vehicle-mounted robot, and the binocular camera is used to shoot a top view or an oblique shot ganoderma lucidum image of the ganoderma lucidum to be detected in the current area (row or group), as shown in fig. 2, where the image acquisition system and the suction device may be separately provided or may be integrally provided. It should be noted that, this shooting angle may cause the ganoderma lucidum to be blocked, so that a part of ganoderma lucidum and other ganoderma lucidum may be blocked in the image.
It should be noted that, because the working environment of the vehicle-mounted robot in this embodiment is the earth ground, the crawler-type travelling mechanism is adopted, and because the crawler-type travelling mechanism is in the prior art, the description is not repeated here.
The sensor system comprises a temperature sensor and a humidity sensor, and is used for measuring temperature information and humidity information of the environment where the ganoderma lucidum area to be detected is located.
The main controller adopts a CPU processor to receive the image of the ganoderma lucidum area to be detected acquired by the image acquisition system and the temperature information and the humidity information acquired by the sensor system, then the main controller performs image processing on the received image of the ganoderma lucidum area to be detected, so that mature ganoderma lucidum is identified, and the main controller sends an identification result to the PID regulator to drive the motor driver, so that the work of the suction device is realized, and the automatic positioning of ganoderma lucidum spore collection by the vehicle-mounted robot is realized.
Specifically, fig. 3 shows a specific implementation process of the master controller for identifying and locating mature ganoderma lucidum, which includes the following specific steps:
firstly, acquiring gray level images and temperature information and humidity information of a ganoderma lucidum area to be detected;
and performing image segmentation on the gray level image to obtain a single ganoderma lucidum image.
In the embodiment, gray-scale images are obtained by carrying out gray-scale treatment on the ganoderma lucidum images to be detected, which are acquired by the image acquisition system of the vehicle-mounted robot; and identifying the single ganoderma lucidum in the gray level image by adopting the semantic segmentation neural network model, so as to obtain the single ganoderma lucidum image.
The semantic segmentation neural network model in the above embodiment may be a full convolution neural network, or may be a U-Net (encoder-decoder); since the semantic segmentation neural network model adopts the existing known network model, the description is not repeated here.
The training process of the semantic segmentation network model in this embodiment includes:
1. building a training data set:
in the embodiment, 500 ganoderma lucidum image data are collected, and ganoderma lucidum image data with 80 percent of the ganoderma lucidum image data are taken as a training data set; the remaining 20% of ganoderma lucidum image data was used as the test set. The collected ganoderma lucidum image data can be images collected in real time or image data collected in history.
2. Tag data:
in this embodiment, taking ganoderma lucidum as an example, since the edge of the fungus cover of ganoderma lucidum and the area of the fungus cover center have obvious color differences, that is, the edge of the fungus cover is generally white, and the area of the fungus cover center is biased to black or brown, different labels can be set during training, wherein the area of the fungus cover center is set to 1, the edge of the fungus cover is set to 2, and the background area is set to 0; the cap center region is a region in the cap edge except the cap edge.
Of course, as other embodiments, the label set during training may also be that the whole bacteria cover area is set to 1, and the background area is set to 0.
3. A loss function is defined. A suitable loss function, such as a cross entropy loss function, is selected.
4. The network is trained. The training data set is input to train the semantic segmentation neural network model, after each round of training, errors are calculated according to the loss function, parameters in the network are adjusted through a back propagation algorithm, and the weight and bias of the network are optimized, so that the performance of the network is improved.
In the embodiment, the trained semantic segmentation network model is utilized to segment gray level images, so that the edge and the heart area of the fungus cover of the ganoderma lucidum can be identified; wherein the fungus cover heart area is a single ganoderma lucidum image.
Further, because part of ganoderma lucidum in the image may overlap and block each other, when the image is segmented, ganoderma lucidum overlapping together may exist, so that the ganoderma lucidum overlapping together is also needed to be segmented, and the specific process is as follows: because the blocking exists, the areas of the communicating areas between the lucid ganoderma are different, so that whether the blocking exists or not is judged through the area of the communicating areas and a standard area threshold value, wherein the standard area threshold value is the size of a standard image; then, as the labels of the edge of the fungus cover and the heart area of the fungus cover are different, the colors of the various label areas after semantic segmentation and identification are also different, at the moment, according to the one-to-one correspondence between the edge of the fungus cover and the heart area of the fungus cover, the connected domains with the labels of 1 can be directly counted, and the connected domain with each label of 1 is used as one heart area of the fungus cover, so that the single division of the ganoderma lucidum image is shielded.
It should be noted that, the area of the fungus cover heart divided at this time is not a complete area, but is only a part of the area of the fungus cover heart; in this embodiment, the area of the fungus cover heart of the partial area is still used as a single ganoderma lucidum image, and whether the integrity of the area is concerned is not concerned.
In this embodiment, the ambient temperature information and the humidity information of the location where the ganoderma lucidum is located are also required to be obtained. Generally, the temperature and humidity information of all ganoderma lucidum located in the same greenhouse is the same, but there may be differences in temperature and humidity between different local areas, so in order to analyze the maturing condition of ganoderma lucidum more accurately, in this embodiment, the collected temperature and humidity information of ganoderma lucidum area to be detected is used as the temperature and humidity information of all individual ganoderma lucidum for subsequent data analysis.
And secondly, obtaining the average gray level of all pixel points in the single ganoderma lucidum image and the texture similarity of the single ganoderma lucidum image.
It should be noted that, because the same greenhouse/cultivation area is basically the same environment, the overall growth conditions of ganoderma lucidum are similar, and generally there is no great difference. And counting the total pixel gray level value of the identified single ganoderma lucidum image, and calculating the average gray level of the single ganoderma lucidum image according to the number of pixel points in the single ganoderma lucidum image.
As shown in fig. 4, the process of obtaining the texture similarity in this embodiment is:
acquiring a gray level co-occurrence matrix of a single ganoderma lucidum image and a gray level co-occurrence matrix of a standard image;
randomly selecting at least three characteristic quantities in each gray level co-occurrence matrix to construct texture characteristics of a single ganoderma lucidum image and a standard image;
and calculating the texture similarity of the texture features of the single ganoderma lucidum image and the texture features of the standard image.
Specifically, since the gray level co-occurrence matrix is mainly used for extracting the texture features of the image, and involves 14 feature quantities, the information of the image can be represented by the gray level co-occurrence matrix; in this embodiment, only four representative and common statistics are selected, which are respectively angular second moment, entropy, contrast and contrast score matrix. Wherein:
the angular second moment is also called energy, is a measurement value of the gray level distribution uniformity degree and the texture thickness of an image, entropy measures the randomness of the image, represents the complexity degree of the image, the contrast reflects the definition degree of the image and the groove depth of the texture, and the contrast matrix reflects the definition degree and the rule degree of the texture.
In calculating the texture similarity, in this embodiment, texture features are formed by counting four feature quantities of the gray level co-occurrence matrix of each single ganoderma lucidum image; and calculating the cosine similarity of texture features corresponding to the standard image, and finally obtaining the texture similarity of all the single ganoderma lucidum images.
Of course, as other embodiments, the present invention may also select three statistics or five statistics therein to obtain the texture similarity.
The standard image in the above steps is obtained by identifying and judging each ganoderma lucidum, such as which ganoderma lucidum is mature, by a professional culturing person, and fusing the mature ganoderma lucidum images.
In this embodiment, the growth condition of the ganoderma lucidum image is represented by obtaining the texture characteristics of the ganoderma lucidum image, and the growth condition is compared with the standard image, so that the difference between the current ganoderma lucidum image and the standard image can be more accurately estimated, and the subsequent maturity condition can be conveniently determined.
Further, since the image capturing system captures images, the single ganoderma lucidum image is not the whole ganoderma lucidum area, when the texture similarity is calculated, the standard ganoderma lucidum image is used as a template to carry out affine transformation on the single ganoderma lucidum image, the maximum overlapping area between each single ganoderma lucidum image and the standard image after affine transformation is obtained, and finally the texture similarity of the texture feature of the large overlapping area corresponding to each single ganoderma lucidum image corresponding to the maximum overlapping area and the texture feature of the large overlapping area corresponding to the standard image is calculated.
In the above steps, by performing region matching of the maximum overlapping region between the single ganoderma lucidum image and the standard image, the condition that the region corresponding to the single ganoderma lucidum image is incomplete, and the region of the standard image is a complete ganoderma lucidum region can be avoided, and when matching is performed, the obtained data have large difference due to incomplete correspondence of the regions, so that the calculation error of the similarity is large, therefore, the region corresponding to the single ganoderma lucidum image needs to be matched in the standard image, the extraction of texture features is performed on the region, and matching of two image data is performed.
According to the invention, the image information of the ganoderma lucidum image can be represented by obtaining the texture similarity and the average gray level, and the growth condition of the ganoderma lucidum can be reflected by the image information, so that data support is provided for the subsequent identification of whether the ganoderma lucidum image is mature or not.
Then, constructing feature vectors based on the environmental temperature information, the humidity information, the average gray level and the texture similarity of the single ganoderma lucidum image to obtain feature vector groups of all the single ganoderma lucidum images; according to the feature vector group, calculating information gain corresponding to any feature of the environment temperature information, the humidity information, the average gray level and the texture similarity in the feature vector group, and calculating to obtain the maturity according to each information gain and the corresponding feature.
Because each individual ganoderma lucidum image corresponds to a group of feature vectors, namely, ambient temperature information, humidity information, average gray scale and texture similarity, in this embodiment, all feature vectors corresponding to the individual ganoderma lucidum images need to be counted, so as to obtain a feature vector group, and any feature of the ambient temperature information, the humidity information, the average gray scale and the texture similarity in the feature vector group is calculated to obtain information gain, and the obtained information gain is used as the weight of the corresponding feature.
The information gain is an important index for selecting the characteristics in the decision tree algorithm, and is defined as how much information can be brought to the classification system by one characteristic, namely, the more information is brought, the more important the characteristic is, and the larger the corresponding information gain is. In other words, the information gain represents the degree to which the complexity (uncertainty) of the information is reduced under one condition.
Where information gain = information entropy-conditional entropy. Since the calculation of the information gain is a well-known technique, the details are not repeated here.
So far, after the information gain of each feature is obtained, four features can be normalized first and then weighted and fused to obtain a fused value which is used as the maturity of the ganoderma lucidum to be detected; by fusing the data in multiple dimension directions into one value, the subsequent calculation amount and the complexity of the algorithm are reduced.
The normalization processing method comprises a min-max normalization method, an atan function conversion method and the like.
When the maturity is calculated, the maturity of all the ganoderma lucidum in the ganoderma lucidum region to be detected can be represented by aiming at the maturity of the ganoderma lucidum region to be detected, namely, starting from the whole region, namely, the maturity, namely, the specific acquisition process is as follows: calculating the ambient temperature information mean value, the humidity information mean value, the average gray level mean value and the texture similarity mean value of all the single ganoderma lucidum images; and carrying out weighted summation on the information gains and corresponding characteristics in the ambient temperature information mean value, the humidity information mean value, the average gray level mean value and the texture similarity mean value to obtain the maturity of the ganoderma lucidum region to be detected.
Of course, in order to evaluate the maturity of ganoderma lucidum more accurately, the maturity of a single ganoderma lucidum can also be calculated, and the specific acquisition process is as follows: and carrying out weighted summation on the information gains and corresponding characteristics in the environmental temperature information, the humidity information, the average gray level and the texture similarity in the single ganoderma lucidum image to obtain the maturity of the single ganoderma lucidum. It should be noted that, in the above description, the maturity of a single ganoderma lucidum image is evaluated, and when the subsequent ganoderma lucidum spores are collected, fine adjustment of the suction device is needed to be performed for positioning; and the maturity of the whole ganoderma lucidum area to be detected is evaluated, fine adjustment of a suction device is not needed, and ganoderma lucidum spores are directly collected.
Finally, comparing the maturity with a standard maturity obtained in advance, and collecting ganoderma spores when the maturity is larger than the standard maturity, wherein ganoderma in a ganoderma area to be detected is mature ganoderma; wherein, the pre-acquired standard maturity is the product of the ratio of the weight of the ganoderma lucidum spores of the historical single mature ganoderma lucidum to the average mature weight and the maturity of the acquired standard image.
The pre-acquired standard maturity in this example is the product of the average maturity weight and the maturity of the acquired standard image by the weight of the ganoderma lucidum spores of the historical single mature ganoderma lucidum.
The ratio is that the average spore weight corresponding to each mature ganoderma lucidum is obtained through the historical ganoderma lucidum spore weight of the single mature ganoderma lucidum counted in advance, and the condition that the ganoderma lucidum spore weight corresponding to the single ganoderma lucidum under normal conditions is more or less can be judged to be in a mature state is obtained according to the experience value of professional breeders. For example, the average spore weight is A, and the spore weight corresponding to the individual ganoderma lucidum can be actually judged to be mature when B is carried out, A is greater than B, and the ratio of B/A is taken as the coefficient for adjusting the maturity of the standard chart image. As one embodiment, the ratio in this example takes a value of 0.8. That is, when the maturity of a single ganoderma lucidum image is greater than the product of the maturity of a standard image and 0.8, the single ganoderma lucidum image belongs to a mature ganoderma lucidum.
In this embodiment, the average spore weight a is generally about 2g, and the individual spore weight is 1.6 g.
So far, after judging the ripe ganoderma lucidum, the vehicle-mounted robot is used for adjusting the suction device to suck the area through the acquired position information of the ripe ganoderma lucidum, so as to collect ganoderma lucidum spores. The position information is obtained by depth information collected by a binocular camera and the position of the center of the connected domain of the mature ganoderma lucidum.
According to the scheme, the mature condition of the ganoderma lucidum is estimated by respectively starting from two directions of image information of the ganoderma lucidum to be picked and environment information of the ganoderma lucidum, namely, on one hand, the growth condition of the ganoderma lucidum and on the other hand, the environment factors required by the growth of the ganoderma lucidum are combined, namely, firstly, the growth condition of the ganoderma lucidum is determined by two factors of gray information and texture information in the image information; secondly, representing the maturity of the ganoderma lucidum from two different dimension directions of the image and the environmental factors; therefore, the accuracy of detecting the maturity of the ganoderma lucidum is improved through comprehensive analysis of the ganoderma lucidum in multiple dimension directions; then, the invention further introduces the factor of ganoderma lucidum spore weight for adjusting the maturity of the standard image, namely determining whether the ganoderma lucidum to be detected is mature or not according to the result after the ganoderma lucidum grows, further ensuring the accuracy and reliability of the detection result, and realizing the accurate positioning of the ganoderma lucidum and the collection of ganoderma lucidum spores.
Fig. 5 is a schematic view schematically showing an automatic positioning device of an in-vehicle picking robot according to an embodiment of the present invention.
In another aspect of the present invention, there is also provided an automatic positioning device of a vehicle-mounted picking robot, including: a processor; and a memory storing computer instructions for in-vehicle picking robot auto positioning that, when executed by the processor, cause the apparatus to perform the in-vehicle picking robot auto positioning method according to one or more embodiments described above.
As shown in fig. 5, the device 501 in the system may include a CPU5011, which may be a general purpose CPU, a special purpose CPU, or other unit of execution for information processing and program execution. Further, the device 501 may further include a mass storage 5012 and a read only memory ROM 5013, wherein the mass storage 5012 may be configured to store various types of data and various programs required for a multimedia network, and the ROM 5013 may be configured to store data required for power-on self test of the device 501, initialization of various functional blocks in the system, driving programs for basic input/output of the system, and booting the operating system.
Further, the device 501 also includes other hardware platforms or components, such as the illustrated TPU (Tensor Processing Unit ) 5014, GPU (Graphic Processing Unit, graphics processor) 5015, FPGAs (Field Programmable Gate Array, field programmable gate arrays) 5016 and MLU (Memory Logic Unit), memory logic unit) 5017. It will be appreciated that while various hardware platforms or components are shown in device 501, this is by way of example only and not limitation, and that one of skill in the art may add or remove corresponding hardware as desired. For example, device 501 may include only a CPU as a well-known hardware platform and another hardware platform as a test hardware platform of the present invention.
The device 501 of the present invention further comprises a communication interface 5018 whereby it is possible to connect to a local area network/wireless local area network (LAN/WLAN) 505 through the communication interface 5018 and further to a local server 506 or to the Internet ("Internet") 507 through the LAN/WLAN. Alternatively or additionally, the device 501 of the present invention may also be directly connected to the internet or cellular network via the communication interface 5018 based on wireless communication technologies, such as third generation ("3G"), fourth generation ("4G"), or 5 th generation ("5G") wireless communication technologies. In some application scenarios, the device 501 of the present invention may also access a server 508 and possibly a database 509 of an external network as needed.
The peripheral devices of the device 501 may include a display means 502, an input means 503 and a data transmission interface 504. In one embodiment, the display device 502 may include, for example, one or more speakers and/or one or more visual displays. The input device 503 may include, for example, a keyboard, mouse, microphone, gesture-capturing camera, or other input buttons or controls configured to receive input of data or user instructions. The data transfer interface 504 may include, for example, a serial interface, a parallel interface, or a universal serial bus interface ("USB"), a small computer system interface ("SCSI"), serial ATA, fireWire ("FireWire"), PCI Express, and high definition multimedia interface ("HDMI"), etc., configured for data transfer and interaction with other devices or systems.
The above-described CPU5011, mass memory 5012, read only memory ROM 5013, TPU 5014, GPU 5015, FPGA 5016, MLU 5017, and communication interface 5018 of the device 501 of the present invention can be connected to each other through a bus 5019, and data interaction with peripheral devices can be achieved through the bus. In one embodiment, through the bus 5019, the cpu5011 may control other hardware components in the device 501 and its peripherals.
In operation, the processor CPU5011 of the apparatus 501 of the present invention may obtain media data packets via the input device 503 or the data transfer interface 504 and retrieve computer program instructions or code stored in the memory 5012 to control the automated positioning of the in-vehicle picking robot.
From the above description of the modular design of the present invention, it can be seen that the system of the present invention can be flexibly arranged according to the application scenario or requirement and is not limited to the architecture shown in the drawings. Further, it should also be appreciated that any module, unit, component, server, computer, or device that performs the operations of the examples of the invention may include or otherwise access a computer-readable medium, such as a storage medium, a computer storage medium, or a data storage device (removable) and/or non-removable) such as, for example, a magnetic disk, optical disk, or magnetic tape. Computer storage media may include volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data.
Based on this, the invention also discloses a computer readable storage medium, on which is stored computer readable instructions for automatic positioning of a vehicle-mounted picking robot, which when executed by one or more processors, implement the vehicle-mounted picking robot automatic positioning method as described in one or more embodiments above.
While various embodiments of the present invention have been shown and described herein, it will be obvious to those skilled in the art that such embodiments are provided by way of example only. Many modifications, changes, and substitutions will now occur to those skilled in the art without departing from the spirit and scope of the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in practicing the invention.
Claims (6)
1. An automatic positioning method of a vehicle-mounted picking robot is characterized by comprising the following steps:
acquiring gray level images, ambient temperature information and humidity information of a ganoderma lucidum area to be detected; image segmentation is carried out on the gray level image to obtain a single ganoderma lucidum image;
obtaining the average gray level of all pixel points in the single ganoderma lucidum image and the texture similarity of the single ganoderma lucidum image;
constructing feature vectors based on the environmental temperature information, the humidity information, the average gray level and the texture similarity of the single ganoderma lucidum image to obtain feature vector groups of all the single ganoderma lucidum images; according to the feature vector group, calculating information gain corresponding to any feature of the environmental temperature information, the humidity information, the average gray level and the texture similarity in the feature vector group, and calculating to obtain maturity according to each information gain and the corresponding feature;
comparing the maturity with a standard maturity obtained in advance, and collecting ganoderma spores when the maturity is larger than the standard maturity, wherein ganoderma in a ganoderma area to be detected is mature ganoderma; wherein, the pre-acquired standard maturity is the product of the ratio of the weight of the ganoderma lucidum spores of the single ripened ganoderma lucidum to the average ripened weight and the maturity of the acquired standard image;
when the image segmentation is carried out, the method further comprises the step of judging whether the ganoderma lucidum images are blocked, and if the blocking exists between the ganoderma lucidum images, taking the non-blocked area in the blocked ganoderma lucidum images as a single ganoderma lucidum image;
the texture similarity of the single ganoderma lucidum image comprises the following steps:
carrying out affine transformation on the single ganoderma lucidum image to obtain an affine image;
acquiring a maximum overlapping area between the affine image and the standard image;
calculating the texture similarity of a single ganoderma lucidum image and a standard image in the maximum overlapping area;
the information gain is the difference value between the information entropy and the conditional entropy of any feature in the environment temperature information, the humidity information, the average gray level and the texture similarity;
according to the gain of each information and the corresponding characteristics, the obtaining process of the maturity is calculated as follows:
and carrying out weighted summation on the information gains and corresponding characteristics in the environmental temperature information, the humidity information, the average gray level and the texture similarity in the single ganoderma lucidum image to obtain the maturity of the single ganoderma lucidum.
2. The automatic positioning method of a vehicle-mounted picking robot according to claim 1, wherein the obtaining process of the texture similarity of the single ganoderma lucidum image is as follows:
acquiring gray level co-occurrence matrixes of all single ganoderma lucidum images and standard images;
randomly selecting at least three characteristic quantities in gray level co-occurrence matrixes corresponding to any single ganoderma lucidum image and standard image, and constructing texture characteristics of any single ganoderma lucidum image and standard image;
and calculating the texture similarity of the texture characteristics of any single ganoderma lucidum image and the standard image.
3. The automatic positioning method of a vehicle-mounted picking robot according to claim 2, wherein the characteristic quantity includes at least three of angular second moment, entropy, contrast and contrast score matrix.
4. An automatic positioning method of a vehicle-mounted picking robot is characterized by comprising the following steps:
acquiring gray level images, ambient temperature information and humidity information of a ganoderma lucidum area to be detected; image segmentation is carried out on the gray level image to obtain a single ganoderma lucidum image;
obtaining the average gray level of all pixel points in the single ganoderma lucidum image and the texture similarity of the single ganoderma lucidum image;
constructing feature vectors based on the environmental temperature information, the humidity information, the average gray level and the texture similarity of the single ganoderma lucidum image to obtain feature vector groups of all the single ganoderma lucidum images; according to the feature vector group, calculating information gain corresponding to any feature of the environmental temperature information, the humidity information, the average gray level and the texture similarity in the feature vector group, and calculating to obtain maturity according to each information gain and the corresponding feature;
comparing the maturity with a standard maturity obtained in advance, and collecting ganoderma spores when the maturity is larger than the standard maturity, wherein ganoderma in a ganoderma area to be detected is mature ganoderma; wherein, the pre-acquired standard maturity is the product of the ratio of the weight of the ganoderma lucidum spores of the single ripened ganoderma lucidum to the average ripened weight and the maturity of the acquired standard image;
when the image segmentation is carried out, the method further comprises the step of judging whether the ganoderma lucidum images are blocked, and if the blocking exists between the ganoderma lucidum images, taking the non-blocked area in the blocked ganoderma lucidum images as a single ganoderma lucidum image;
the texture similarity of the single ganoderma lucidum image comprises the following steps:
carrying out affine transformation on the single ganoderma lucidum image to obtain an affine image;
acquiring a maximum overlapping area between the affine image and the standard image;
calculating the texture similarity of a single ganoderma lucidum image and a standard image in the maximum overlapping area;
the information gain is the difference value between the information entropy and the conditional entropy of any feature in the environment temperature information, the humidity information, the average gray level and the texture similarity; according to the gain of each information and the corresponding characteristics, the obtaining process of the maturity is calculated as follows:
calculating the ambient temperature information mean value, the humidity information mean value, the average gray level mean value and the texture similarity mean value of all the single ganoderma lucidum images;
and carrying out weighted summation on the information gains and corresponding characteristics in the ambient temperature information mean value, the humidity information mean value, the average gray level mean value and the texture similarity mean value to obtain the maturity of the ganoderma lucidum region to be detected.
5. An automatic positioning device of a vehicle-mounted picking robot, comprising:
a processor; and
a memory storing computer instructions for automatic positioning of a vehicle-mounted picking robot, which when executed by the processor, cause an apparatus to perform the vehicle-mounted picking robot automatic positioning method according to any of claims 1-4.
6. A computer readable storage medium having stored thereon computer readable instructions for automatic positioning of a vehicle-mounted picking robot, which when executed by one or more processors, implement the vehicle-mounted picking robot automatic positioning method of any of claims 1-4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310875948.0A CN116612191B (en) | 2023-07-18 | 2023-07-18 | Automatic positioning method and device for vehicle-mounted picking robot |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310875948.0A CN116612191B (en) | 2023-07-18 | 2023-07-18 | Automatic positioning method and device for vehicle-mounted picking robot |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116612191A CN116612191A (en) | 2023-08-18 |
| CN116612191B true CN116612191B (en) | 2023-09-22 |
Family
ID=87675026
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310875948.0A Active CN116612191B (en) | 2023-07-18 | 2023-07-18 | Automatic positioning method and device for vehicle-mounted picking robot |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116612191B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109522899A (en) * | 2018-10-16 | 2019-03-26 | 江苏省无线电科学研究所有限公司 | Detection method, device and the electronic equipment of coffee ripening fruits |
| CN111753577A (en) * | 2019-03-28 | 2020-10-09 | 天津工业大学 | Apple identification and positioning method in automatic picking robot |
| CN113298835A (en) * | 2021-06-16 | 2021-08-24 | 大连大学 | Fruit harvesting maturity weighted evaluation method |
| CN114898000A (en) * | 2022-05-06 | 2022-08-12 | 塔里木大学 | Red date mature picking analysis method, system and terminal considering environmental factors |
| CN115223164A (en) * | 2022-08-29 | 2022-10-21 | 广西壮族自治区农业科学院 | Melon maturity detection method and system based on artificial intelligence |
| CN115984847A (en) * | 2022-11-22 | 2023-04-18 | 浙江寿仙谷植物药研究院有限公司 | Ganoderma lucidum spore powder microscopic image identification method and device based on computer vision |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11636672B2 (en) * | 2019-10-15 | 2023-04-25 | A.U.G. Signals Ltd. | Crop phenology estimation and tracking with remote sensing imagery |
| CN114169618B (en) * | 2021-12-11 | 2022-08-09 | 广东工业大学 | Kiwi fruit maturity nondestructive identification prediction method based on machine vision learning |
-
2023
- 2023-07-18 CN CN202310875948.0A patent/CN116612191B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109522899A (en) * | 2018-10-16 | 2019-03-26 | 江苏省无线电科学研究所有限公司 | Detection method, device and the electronic equipment of coffee ripening fruits |
| CN111753577A (en) * | 2019-03-28 | 2020-10-09 | 天津工业大学 | Apple identification and positioning method in automatic picking robot |
| CN113298835A (en) * | 2021-06-16 | 2021-08-24 | 大连大学 | Fruit harvesting maturity weighted evaluation method |
| CN114898000A (en) * | 2022-05-06 | 2022-08-12 | 塔里木大学 | Red date mature picking analysis method, system and terminal considering environmental factors |
| CN115223164A (en) * | 2022-08-29 | 2022-10-21 | 广西壮族自治区农业科学院 | Melon maturity detection method and system based on artificial intelligence |
| CN115984847A (en) * | 2022-11-22 | 2023-04-18 | 浙江寿仙谷植物药研究院有限公司 | Ganoderma lucidum spore powder microscopic image identification method and device based on computer vision |
Non-Patent Citations (4)
| Title |
|---|
| 基于混合特征与最小二乘支持向量机的山核桃鲜果识别;章云;周竹;周素茵;;中南林业科技大学学报(第11期);全文 * |
| 温室绿熟番茄机器视觉检测方法;李寒;张漫;高宇;李民赞;季宇寒;;农业工程学报(第S1期);全文 * |
| 灵芝图像分割技术研究;袁雪霞;尹新富;;中国食用菌(第03期);全文 * |
| 苹果采摘机器人夜间图像边缘保持的Retinex增强算法;姬伟;吕兴琴;赵德安;贾伟宽;丁世宏;;农业工程学报(第06期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116612191A (en) | 2023-08-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109785337B (en) | In-column mammal counting method based on example segmentation algorithm | |
| CN110060233B (en) | Corn ear damage detection method | |
| Lin et al. | The pest and disease identification in the growth of sweet peppers using faster R-CNN and mask R-CNN | |
| CN107292228A (en) | A kind of method for accelerating face recognition search speed | |
| Rahim et al. | Deep learning-based accurate grapevine inflorescence and flower quantification in unstructured vineyard images acquired using a mobile sensing platform | |
| CN115661650A (en) | Farm management system based on data monitoring of Internet of things | |
| CN112580671A (en) | Automatic detection method and system for multiple development stages of rice ears based on deep learning | |
| Zermas et al. | Extracting phenotypic characteristics of corn crops through the use of reconstructed 3D models | |
| CN111783751A (en) | Rifle ball linkage and BIM-based breeding house piglet abnormity early warning method | |
| CN115294518A (en) | Intelligent monitoring method and system for accurate cultivation of horticultural plants in greenhouse | |
| CN113822198B (en) | Peanut growth monitoring method, system and medium based on UAV-RGB image and deep learning | |
| CN109166127B (en) | Wearable plant phenotype sensing system | |
| CN113793385A (en) | Method and device for positioning fish head and fish tail | |
| CN116612191B (en) | Automatic positioning method and device for vehicle-mounted picking robot | |
| KR102393265B1 (en) | System for detecting pests of shiitake mushrooms | |
| CN111369497B (en) | Walking type tree fruit continuous counting method and device | |
| CN116739739A (en) | Loan amount evaluation method and device, electronic equipment and storage medium | |
| CN114700941B (en) | Strawberry picking method based on binocular vision and robot system | |
| CN112329697B (en) | Improved YOLOv 3-based on-tree fruit identification method | |
| Fang et al. | Classification system study of soybean leaf disease based on deep learning | |
| CN115294472A (en) | Fruit yield estimation method, model training method, equipment and storage medium | |
| WO2022137242A1 (en) | Deep learning system for estimating object volume, weight and crop yields based on remote imaging | |
| CN115830474A (en) | Method and system for identifying wild Tibetan medicine lamiophlomis rotata and distribution thereof and calculating yield thereof | |
| CN112488230A (en) | Crop water stress degree judging method and device based on machine learning | |
| James et al. | CitDet: A Benchmark Dataset for Citrus Fruit Detection |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |