CN111242093A - Cordyceps sinensis field target identification equipment and method - Google Patents
Cordyceps sinensis field target identification equipment and method Download PDFInfo
- Publication number
- CN111242093A CN111242093A CN202010116762.3A CN202010116762A CN111242093A CN 111242093 A CN111242093 A CN 111242093A CN 202010116762 A CN202010116762 A CN 202010116762A CN 111242093 A CN111242093 A CN 111242093A
- Authority
- CN
- China
- Prior art keywords
- image
- cordyceps sinensis
- target
- module
- imaging device
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F16—ENGINEERING ELEMENTS AND UNITS; GENERAL MEASURES FOR PRODUCING AND MAINTAINING EFFECTIVE FUNCTIONING OF MACHINES OR INSTALLATIONS; THERMAL INSULATION IN GENERAL
- F16M—FRAMES, CASINGS OR BEDS OF ENGINES, MACHINES OR APPARATUS, NOT SPECIFIC TO ENGINES, MACHINES OR APPARATUS PROVIDED FOR ELSEWHERE; STANDS; SUPPORTS
- F16M13/00—Other supports for positioning apparatus or articles; Means for steadying hand-held apparatus or articles
- F16M13/04—Other supports for positioning apparatus or articles; Means for steadying hand-held apparatus or articles for supporting on, or holding steady relative to, a person, e.g. by chains, e.g. rifle butt or pistol grip supports, supports attached to the chest or head
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T1/00—General purpose image data processing
- G06T1/20—Processor architectures; Processor configuration, e.g. pipelining
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
- G08B21/18—Status alarms
- G08B21/24—Reminder alarms, e.g. anti-loss alarms
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- General Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Emergency Management (AREA)
- Mechanical Engineering (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the field of intelligent agricultural equipment technology development and utilization, and discloses cordyceps sinensis field target recognition equipment and a method. The invention can realize the purpose of conveniently identifying the cordyceps sinensis in the field, and the identification equipment is used for searching the target in the field, thereby improving the efficiency of searching the cordyceps sinensis in the field and being more beneficial to searching the cordyceps sinensis.
Description
Technical Field
The invention relates to the field of intelligent agricultural equipment technology development and utilization, in particular to a device and a method for identifying a cordyceps sinensis field target.
Background
The cordyceps sinensis is a unique and precious medicinal material in China, and has multiple effects of regulating immune system, resisting tumor, resisting fatigue and the like. The cordyceps sinensis is a specific bacterium substance in the Qinghai-Tibet plateau, is limited in distribution region, is concentrated in a region with the altitude of more than 3000 meters, is distributed in Qilian mountains in northern China, southern China to northern mountain in Yunnan China, east China is from plateau regions in western Sichuan, and most regions of Xida Himalaya, including provinces of Tibet, Qinghai, Sichuan, Yunnan, Gansu and the like.
However, the collection of cordyceps sinensis is mainly carried out in a pure manual mode, and a large amount of farmers and workers go to the production area to be collected and dug every year in four to six months. However, because the cordyceps sinensis grows sparsely, the grass bodies exposed out of the ground are small, the grown surface weeds are complex, and the mining process is just like a sea fishing needle. During the search, the digger needs to kneel or crawl over a wet mountain, slowly search through eye force, and then dig. Such conventional mining methods have three serious problems as follows: first, the pure manual mining method is slow and inefficient, and usually a skilled miner can only mine dozens of cordyceps every day, thus requiring a large amount of manpower to mine. Secondly, the wormhole for growing the cordyceps is moist, the plateau is strong in ultraviolet rays, and the long-time creeping searching process has great damage to the body of the digger. Thirdly, a large amount of manual residence and large-scale activities on meadows also bring great influence on the bodies of the grasses and the environment, and damage the natural ecology of the plateau areas.
Disclosure of Invention
Based on the problems, the invention provides cordyceps sinensis field target identification equipment and a method. The invention can realize the purpose of conveniently identifying the cordyceps sinensis in the field, and the identification equipment is used for searching the target in the field, thereby improving the efficiency of searching the cordyceps sinensis in the field and being more beneficial to searching the cordyceps sinensis.
In order to solve the technical problems, the technical scheme adopted by the invention is as follows:
the imaging device is electrically connected with the core control and calculation module through a transmission line, and the core control and calculation module is connected with an alarm module for warning the existence of the cordyceps sinensis and a power supply module for providing power for the device.
As a preferred mode, the core control and calculation module comprises a circuit main board, a calculation acceleration module, a calculation chip and a storage device, wherein the calculation chip comprises a first-stage target detection module for outputting one or more candidate areas with cordyceps sinensis targets and a second-stage area target re-judgment module for judging whether the scene image has the cordyceps sinensis targets, the first-stage target detection module comprises a target detector for detecting the cordyceps sinensis targets, and the second-stage area target re-judgment module comprises a judger for judging the cordyceps sinensis targets; the alarm module is connected with the computing chip, and the power supply module is connected with the circuit main board.
As a preferable mode, the alarm module is composed of one or more of a buzzer and an indicator light, and the power supply module is composed of one or more of a storage battery and a solar charging panel.
Preferably, the imaging device is provided with a fixing device for fixing the imaging device, and the core control and calculation module is arranged in the fixing device.
As a preferable mode, the image forming apparatus is provided with a stabilizing device for reducing shaking of the image forming apparatus.
Preferably, the imaging device is provided with a measuring bracket for keeping the imaging device at a constant angle and distance from the ground surface.
A method for identifying wild targets of Cordyceps sinensis comprises the following steps:
s01, holding the imaging device to continuously scan the earth surface, and acquiring a scene image by an imaging system of the imaging device in real time in a view range;
s02: the imaging equipment transmits the acquired scene image to the core control and calculation module through a transmission line;
s03: the core control and calculation module performs detection analysis on the scene image, and if the detected scene image contains cordyceps sinensis, the S04 step is performed; otherwise, continuing to process the next scene image obtained from the step of S01;
s04: and if the scene image contains the cordyceps sinensis target, triggering the warning module.
As a preferred mode, the detection process of the core control and calculation module is as follows: the method comprises the following steps:
s201: after receiving a scene image, performing conventional image preprocessing on the scene image and outputting a preprocessed image;
s202, outputting a low-resolution image after the resolution of the preprocessed image is reduced;
s203, outputting one or more target candidate areas with cordyceps sinensis to the low-resolution image by using an image target detection algorithm;
s204, intercepting a high-resolution image corresponding to the target candidate frame from the preprocessed image according to the target candidate area;
and S205, judging whether the cordyceps sinensis exists according to the high-resolution image.
In a preferred mode, the method for identifying the wild target of the cordyceps sinensis comprises a training method for identifying the cordyceps sinensis: the method comprises the following steps:
s301: collecting an initial image containing Cordyceps sinensis;
s302: calibrating coordinates of two top points of a rectangular area where the cordyceps sinensis is located in the initial image, and outputting a primary processing image;
s303: performing image processing on the primary processing image to generate a secondary processing image, calculating the calibration coordinate information of the cordyceps sinensis in the secondary processing image, and outputting the secondary processing image and the calibration coordinate information of the cordyceps sinensis in the secondary processing image;
s304: adjusting the image resolution of the secondary processing image, outputting a low-resolution secondary processing image, recalculating the calibration coordinate information of the cordyceps sinensis according to the resolution adjustment ratio, and outputting the calibration coordinate information;
s305: training a first-stage target detection module by using the low-resolution secondary processing image and the calibration coordinate information to obtain target detector parameters;
s306: according to the secondary processing image and the calibration coordinate information of the cordyceps sinensis in the secondary processing image, cutting a rectangular image area where the cordyceps sinensis is located from the secondary processing image, performing data enhancement to construct a positive sample, randomly cutting three times of background weed rectangular image areas of the positive sample in target number from the secondary processing image to construct a negative sample, and scaling the rectangular areas of the positive sample and the negative sample to the same size;
s307: and training a second-stage area target re-judgment module by utilizing positive and negative samples to obtain a judger parameter.
As a preferred mode, the first-stage target detection module employs a deep neural network, and the second-stage regional target re-determination module employs a convolutional neural network.
Compared with the prior art, the invention has the beneficial effects that:
(1) according to the invention, the acquired ground surface video image is transmitted to the core control and calculation module for analysis through the imaging equipment, and the alarm device is used for alarming the cordyceps sinensis in the video image, so that the cordyceps sinensis can be found quickly and effectively, and the difficulty of manually searching the cordyceps sinensis is reduced;
(2) the invention reduces the shake of the imaging equipment in the using process through the stabilizing device, enhances the stability of the imaging equipment and is beneficial to making video images clearer;
(3) the angle and the distance between the imaging equipment and the earth surface are kept through the measuring bracket, and the measuring bracket is in contact with the earth surface, so that the range of the imaging visual field is favorably determined;
drawings
FIG. 1 is a schematic structural diagram of the present invention.
FIG. 2 is a flow chart of the detection of the core control and computation module.
FIG. 3 is a flowchart of a training method for the first stage target detection module and the second stage target detection module.
The system comprises an imaging device 1, a stabilizing device 2, a fixing device 3, a measuring support 4, a transmission line 5, a core control and calculation module 6, an alarm module 7, a power supply module 8, an image preprocessing module 601, an image resolution adjusting module 602, a first-stage target detection module 603, a candidate high-resolution image module 604, a second-stage regional target re-judgment module 605, an artificial calibration training data module 701, a data amplification module 702, an image resolution adjusting module 703 and a high-resolution target and background module 704.
Detailed Description
The invention will be further described with reference to the accompanying drawings. Embodiments of the present invention include, but are not limited to, the following examples.
Example 1:
referring to fig. 1, the cordyceps sinensis field target identification device comprises an imaging device 1 for collecting ground surface scene images and a core control and calculation module 6 for controlling all parts and analyzing the scene images, wherein the imaging device 1 is electrically connected with the core control and calculation module 6 through a transmission line 5, and the core control and calculation module 6 is connected with an alarm module 7 for warning the existence of cordyceps sinensis and a power supply module 8 for supplying power to the device.
In this embodiment, the imaging device 1 is used for collecting surface video images, the surface video images are transmitted to the core control and calculation module 6 through the transmission line 5, the core control and calculation module 6 is used for controlling the electronic control of each component and calculating and analyzing the video data, so as to analyze whether the cordyceps sinensis exists in the imaging field of the imaging device 1, when the cordyceps sinensis exists in the imaging field and is detected by the core control and calculation module 6, the alarm module 7 sends out warning information, and the power supply module 8 is used for supplying power to the imaging device 1, the core control and calculation module 6 and the alarm module 7, so as to reduce the difficulty of manually searching the cordyceps sinensis.
The core control and calculation module 6 comprises a circuit main board, a calculation acceleration module, a calculation chip and a storage device, wherein the calculation chip comprises a first-stage target detection module 603 for outputting one or more candidate areas with cordyceps sinensis targets and a second-stage area target re-judgment module 605 for judging whether the images have the cordyceps sinensis targets or not according to the images, the first-stage target detection module 603 comprises a target detector for detecting the cordyceps sinensis targets, and the second-stage area target re-judgment module 605 comprises a judger for judging the cordyceps sinensis targets; the alarm module 7 is connected with the computing chip, and the power supply module 8 is connected with the circuit main board.
The circuit main board adopts an Up Squared board or a raspberry group, the calculation acceleration module adopts a calculation acceleration chip, the calculation acceleration chip compresses model parameters trained in the core control and calculation module, and then loads the compressed model and parameters into a Movidius hardware, the Movidius hardware is provided with a special chip VPU (vector Processing Unit) which can accelerate the deduction speed of the core control and calculation module for target detection, the calculation chip adopts a CPU N4200 or a BCM2837, and the storage device adopts a 16G memory, a 256G hard disk or a 64GTF card. The alarm module 7 is composed of one or more of a buzzer and an indicator light, and the power supply module 8 is composed of one or more of a storage battery and a solar charging panel. The power supply module 8 supplies power to the imaging device 1, the computing chip, the storage device, the alarm module 7, and the like through the transmission line 5.
The imaging device 1 is provided with a fixing device 3 for fixing the imaging device 1, and the core control and calculation module 6 is arranged in the fixing device 3; the fixing device 3 is any one of a handheld rod, an unmanned aerial vehicle and a robot dog. The fixing device 3 is used for fixing the imaging device 1 and protecting the core control and calculation module 6. Any one of specific optional handheld pole, unmanned aerial vehicle and the machine dog of using, handheld pole can make the operator make imaging device 1 press close to with ground under the circumstances of not crooked back, and unmanned aerial vehicle can reduce the operator and lead to the fact the destruction to the meadow when seeking the chinese caterpillar fungus, and the machine dog can make the operator need not hand imaging device 1, and is fast at mountain meadow walking, is favorable to accelerating the search to the chinese caterpillar fungus.
Be equipped with on imaging device 1 and be used for alleviateing the stabilising arrangement 2 that imaging device 1 rocked, because when searching on the mountain meadow, imaging device 1 produces easily and rocks, and the image quality that leads to formation of image is unclear, and stabilising arrangement 2 is favorable to alleviateing imaging device 1 and rocks, makes the image quality of formation of image clear.
The imaging device 1 is provided with a measuring bracket 4 for keeping the imaging device 1 at a constant angle and distance from the earth surface, when the fixing device 3 for fixing the imaging device 1 is a handheld rod, the fixing measuring bracket 4 needs to be in contact with the earth surface, so that the measuring bracket 4 can keep a certain distance and angle from the earth surface, and the imaging view field of the imaging device 1 can be determined; fixing device 3 when fixed imaging device 1 is unmanned aerial vehicle or machine dog, and unmanned aerial vehicle keeps the horizontal attitude flight all the time, and the machine dog removes on ground all the time, and the fixed bolster then need not contact with ground, only needs adjustment imaging device 1 at fixing device 3's angle, makes imaging device 1 and ground keep certain distance and angle to confirm imaging device 1's the formation of image field of vision.
Example 2:
this example is a specific application of example 1.
Referring to fig. 1, a method for identifying wild targets of cordyceps sinensis comprises the following steps: the method comprises the following steps:
s01, holding the imaging device 1 to continuously scan the earth surface, and acquiring a scene image by an imaging system of the imaging device 1 in real time in a view range;
s02: the imaging device 1 transmits the collected scene image to the core control and calculation module through the transmission line 5;
s03: the core control and calculation module 6 performs detection analysis on the scene image, and if the detected scene image contains cordyceps sinensis, the step S04 is performed; otherwise, continuing to process the next scene image obtained from the step of S01;
s04: and if the scene image contains the cordyceps sinensis target, triggering the warning module.
The imaging system is usually color imaging, but can be expanded to use imaging information such as depth of field, infrared, multispectral, hyperspectral, etc.
Referring to fig. 2, the detection process of the core control and calculation module 6 is as follows: the method comprises the following steps:
s201: after receiving a scene image, performing conventional image preprocessing on the scene image and outputting a preprocessed image;
s202, outputting a low-resolution image after the resolution of the preprocessed image is reduced;
s203, outputting one or more target candidate areas with cordyceps sinensis to the low-resolution image by using an image target detection algorithm;
s204, intercepting a high-resolution image corresponding to the target candidate frame from the preprocessed image according to the target candidate area;
and S205, judging whether the cordyceps sinensis exists according to the high-resolution image.
The image preprocessing module 601 performs scene image preprocessing including, but not limited to, automatic brightness adjustment, color adjustment, and the like, and outputs a command for preprocessing an image; the image resolution provided by the conventional imaging device 1 is high, and the image resolution adjusting module 602 executes a command of reducing the resolution of the preprocessed image and outputting a low-resolution image, so that the high-speed actual measurement of the cordyceps sinensis is facilitated; the first-stage object detection module 603 executes a command for outputting one or more object candidate regions in which cordyceps sinensis exists from the low-resolution image using an object detection algorithm, the object detection determining a position and a category of a specific object in the inputted scene image; the candidate region high resolution image module 604 performs a command to increase the resolution of the target candidate region and outputs a high resolution image, which is higher than the resolution of the low resolution image and may be the same as the resolution of the pre-processed image; the second-stage region target re-determination module 605 executes a command for determining whether the cordyceps sinensis exists in the high-resolution image, and re-determines whether the region is a cordyceps sinensis target.
Referring to fig. 3, a method for identifying a cordyceps sinensis field target includes a training method for identifying cordyceps sinensis: the method comprises the following steps:
s301: collecting an initial image containing Cordyceps sinensis;
s302: calibrating coordinates of two top points of a rectangular area where the cordyceps sinensis is located in the initial image, and outputting a primary processing image;
s303: performing image processing on the primary processing image to generate a secondary processing image, calculating the calibration coordinate information of the cordyceps sinensis in the secondary processing image, and outputting the secondary processing image and the calibration coordinate information of the cordyceps sinensis in the secondary processing image;
s304: adjusting the image resolution of the secondary processing image, outputting a low-resolution secondary processing image, recalculating the calibration coordinate information of the cordyceps sinensis according to the resolution adjustment ratio, and outputting the calibration coordinate information;
s305: training a first-stage target detection module 603 by using the low-resolution secondary processing image and the calibration coordinate information to obtain target detector parameters;
s306: according to the secondary processing image and the calibration coordinate information of the cordyceps sinensis in the secondary processing image, cutting a rectangular image area where the cordyceps sinensis is located from the secondary processing image, performing data enhancement to construct a positive sample, randomly cutting three times of background weed rectangular image areas of the positive sample in target number from the secondary processing image to construct a negative sample, and scaling the rectangular areas of the positive sample and the negative sample to the same size;
s307: the second level region target re-decision module 605 is trained using the positive and negative samples to obtain the decider parameters.
The manual calibration training data module 701 performs coordinate calibration on two vertex coordinates of a rectangular area where the cordyceps sinensis is located in the initial image and outputs a primary processing image, wherein the two vertex coordinates are specifically upper left point coordinates and lower right point coordinates of the rectangular area, and the step is to output the initial processing image for calibrating the upper left point coordinates and the lower right point coordinates of the rectangular area where the cordyceps sinensis is located; the data amplification module 702 executes the commands of calculating the calibrated coordinate information of the cordyceps sinensis after performing image processing on the primary processed image and outputting a secondary processed image and the calibrated coordinate information of the cordyceps sinensis in the secondary processed image, wherein the image processing is the processing of performing mirror image, rotation, scaling, brightness, tone adjustment and the like on the primary processed image; the image resolution adjusting module 602 executes the image resolution of the secondary processing image to be reduced, outputs the secondary processing image with low resolution, recalculates the calibration coordinate information of the cordyceps sinensis according to the adjustment of the resolution, and outputs a command of the calibration coordinate information; the first stage target detection module 603 is trained according to the low resolution secondary processed image and the calibration coordinate information to obtain target detector parameters. The high-resolution target and background module 704 is used for cutting rectangular image areas where the cordyceps sinensis is located in the secondary processing image, performing data enhancement to construct positive samples, randomly cutting three times the number of the rectangular image areas of the background weeds in the secondary processing image to construct negative samples, scaling the rectangular areas of the positive samples and the negative samples to the same size, and constructing commands of positive and negative samples; the second level region target re-decision module 605 is trained according to the positive and negative samples to obtain the parameters of the decision device.
The first level target detection module 603 employs a deep neural network and the second level regional target re-determination module 605 employs a convolutional neural network. The second-stage regional target re-determination module 605 specifically performs a specific design for the second-class classifier based on the convolutional neural network, where the selected deep neural network is formed by multiple layers of convolutional neural networks, and the convolutional kernel size of each layer of convolutional neural network and the total number of network layers need to be determined. The deep neural network adopts a movidius neural computation chip.
The above is an embodiment of the present invention. The embodiments and specific parameters in the embodiments are only used for clearly illustrating the verification process of the invention and are not used for limiting the patent protection scope of the invention, which is defined by the claims, and all the equivalent structural changes made by using the contents of the description and the drawings of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. A cordyceps sinensis field target recognition device is characterized in that: the system comprises an imaging device (1) for acquiring an earth surface scene image and a core control and calculation module (6) for controlling each part and analyzing the scene image, wherein the imaging device (1) is electrically connected with the core control and calculation module (6) through a transmission line (5), and the core control and calculation module (6) is connected with an alarm module (7) for warning the existence of cordyceps sinensis and a power supply module (8) for providing power for the device.
2. The cordyceps sinensis field target recognition device according to claim 1, wherein: the core control and calculation module (6) comprises a circuit main board, a calculation acceleration module, a calculation chip and a storage device, wherein the calculation chip comprises a first-stage target detection module (603) for outputting one or more candidate areas with cordyceps sinensis targets and a second-stage area target re-judgment module (605) for judging whether the cordyceps sinensis exists in a scene image, the first-stage target detection module (603) comprises a target detector for detecting the cordyceps sinensis, and the second-stage area target re-judgment module (605) comprises a judger for judging the cordyceps sinensis; the alarm module (7) is connected with the computing chip, and the power supply module (8) is connected with the circuit main board.
3. The cordyceps sinensis field target recognition device according to claim 1, wherein: the alarm module (7) is composed of one or more of a buzzer and an indicator light, and the power supply module (8) is composed of one or more of a storage battery and a solar charging panel.
4. The cordyceps sinensis field target recognition device according to claim 1, wherein: the imaging device (1) is provided with a fixing device (3) for fixing the imaging device (1), and the core control and calculation module (6) is arranged in the fixing device (3).
5. The cordyceps sinensis field target recognition device according to claim 1, wherein: the imaging device (1) is provided with a stabilizing device (2) for reducing shaking of the imaging device (1).
6. The cordyceps sinensis field target recognition device according to claim 1, wherein: the imaging device (1) is provided with a measuring bracket (4) used for keeping the imaging device (1) and the ground surface to have a constant angle and distance.
7. A method for identifying wild targets of cordyceps sinensis is characterized by comprising the following steps: the method comprises the following steps:
s01, holding the imaging device (1) to continuously scan the earth surface, and acquiring a scene image by an imaging system of the imaging device (1) in real time in a view range;
s02: the imaging device (1) transmits the acquired scene images to a core control and calculation module (6) through a transmission line (5);
s03: the core control and calculation module (6) performs detection analysis on the scene image, and if the detected scene image contains cordyceps sinensis, the step of S04 is performed; otherwise, continuing to process the next scene image obtained from the step of S01;
s04: and if the scene image contains the cordyceps sinensis target, triggering the warning module.
8. The method for identifying the wild target of cordyceps sinensis according to claim 7, wherein the method comprises the following steps: the detection process of the core control and calculation module (6) is as follows: the method comprises the following steps:
s201: after receiving a scene image, performing conventional image preprocessing on the scene image and outputting a preprocessed image;
s202, outputting a low-resolution image after the resolution of the preprocessed image is reduced;
s203, outputting one or more target candidate areas with cordyceps sinensis to the low-resolution image by using an image target detection algorithm;
s204, intercepting a high-resolution image corresponding to the target candidate frame from the preprocessed image according to the target candidate area;
and S205, judging whether the cordyceps sinensis exists according to the high-resolution image.
9. The method for identifying the wild target of cordyceps sinensis according to claim 7, wherein the method comprises the following steps: the training method comprises the following steps of identifying the cordyceps sinensis: the method comprises the following steps:
s301: collecting an initial image containing Cordyceps sinensis;
s302: calibrating coordinates of two top points of a rectangular area where the cordyceps sinensis is located in the initial image, and outputting a primary processing image;
s303: performing image processing on the primary processing image to generate a secondary processing image, calculating the calibration coordinate information of the cordyceps sinensis in the secondary processing image, and outputting the secondary processing image and the calibration coordinate information of the cordyceps sinensis in the secondary processing image;
s304: adjusting the image resolution of the secondary processing image, outputting a low-resolution secondary processing image, recalculating the calibration coordinate information of the cordyceps sinensis according to the resolution adjustment ratio, and outputting the calibration coordinate information;
s305: training a first-stage target detection module (603) by using the low-resolution secondary processing image and the calibration coordinate information to obtain target detector parameters;
s306: according to the secondary processing image and the calibration coordinate information of the cordyceps sinensis in the secondary processing image, cutting a rectangular image area where the cordyceps sinensis is located from the secondary processing image, performing data enhancement to construct a positive sample, randomly cutting three times of background weed rectangular image areas of the positive sample in target number from the secondary processing image to construct a negative sample, and scaling the rectangular areas of the positive sample and the negative sample to the same size;
s307: and training a second-stage area target re-judgment module (605) by utilizing the positive and negative samples to obtain the parameters of the judger.
10. The method for identifying the wild target of cordyceps sinensis according to claim 7, wherein the method comprises the following steps: the first-stage target detection module (603) adopts a deep neural network, and the second-stage regional target re-judgment module (605) adopts a convolutional neural network.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010116762.3A CN111242093A (en) | 2020-02-25 | 2020-02-25 | Cordyceps sinensis field target identification equipment and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010116762.3A CN111242093A (en) | 2020-02-25 | 2020-02-25 | Cordyceps sinensis field target identification equipment and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111242093A true CN111242093A (en) | 2020-06-05 |
Family
ID=70873368
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010116762.3A Pending CN111242093A (en) | 2020-02-25 | 2020-02-25 | Cordyceps sinensis field target identification equipment and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111242093A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112417193A (en) * | 2020-08-21 | 2021-02-26 | 深圳市小樱桃实业有限公司 | Method and system for searching and identifying field cordyceps sinensis |
-
2020
- 2020-02-25 CN CN202010116762.3A patent/CN111242093A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112417193A (en) * | 2020-08-21 | 2021-02-26 | 深圳市小樱桃实业有限公司 | Method and system for searching and identifying field cordyceps sinensis |
| CN112417193B (en) * | 2020-08-21 | 2021-10-19 | 深圳市小樱桃实业有限公司 | Method and system for searching and identifying field cordyceps sinensis |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Feng et al. | A comprehensive review on recent applications of unmanned aerial vehicle remote sensing with various sensors for high-throughput plant phenotyping | |
| CN107392130A (en) | Classification of Multispectral Images method based on threshold adaptive and convolutional neural networks | |
| CN109886155B (en) | Single-plant rice detection and positioning method, system, equipment and medium based on deep learning | |
| David et al. | Vegetation indices and textures in object-based weed detection from UAV imagery | |
| Sun et al. | Characterizing tree species of a tropical wetland in southern china at the individual tree level based on convolutional neural network | |
| Zhang et al. | A novel method for urban area extraction from VIIRS DNB and MODIS NDVI data: A case study of Chinese cities | |
| Alzadjali et al. | Maize tassel detection from UAV imagery using deep learning | |
| Li et al. | Effects of RapidEye imagery’s red-edge band and vegetation indices on land cover classification in an arid region | |
| CN110516723B (en) | Multi-modal foundation cloud picture identification method based on depth tensor fusion | |
| CN112906537B (en) | Crop identification method and system based on convolutional neural network | |
| CN113657326A (en) | Weed detection method based on multi-scale fusion module and feature enhancement | |
| Shen et al. | Biomimetic vision for zoom object detection based on improved vertical grid number YOLO algorithm | |
| Zhou et al. | Peanut planting area change monitoring from remote sensing images based on deep learning | |
| Bouachir et al. | Computer vision system for automatic counting of planting microsites using UAV imagery | |
| Huang et al. | Depth semantic segmentation of tobacco planting areas from unmanned aerial vehicle remote sensing images in plateau mountains | |
| Zhao et al. | An entropy and MRF model-based CNN for large-scale landsat image classification | |
| CN110414438A (en) | High spectrum image recognition methods based on space clustering Information revision | |
| Chunyang et al. | Sea fog detection using U-Net deep learning model based on MODIS data | |
| Lyu et al. | Development of phenotyping system using low altitude UAV imagery and deep learning | |
| Yuan et al. | Sensitivity examination of YOLOv4 regarding test image distortion and training dataset attribute for apple flower bud classification | |
| CN111242093A (en) | Cordyceps sinensis field target identification equipment and method | |
| Kamal et al. | A preliminary study on machine learning and google earth engine for mangrove mapping | |
| CN112084989A (en) | Unmanned aerial vehicle and CNN-based large-range pine wood nematode withered vertical wood intelligent detection method | |
| Chen et al. | Dense greenhouse extraction in high spatial resolution remote sensing imagery | |
| CN211956489U (en) | Open-air target identification equipment of chinese caterpillar fungus |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20230321 Address after: 518000 201, Building 1, Building 14, Zhongjia Mingyuan, Ronghua Road, Buji Street, Longgang District, Shenzhen, Guangdong Province Applicant after: Shenzhen Ruanjin Biotechnology Co.,Ltd. Address before: Room 203, No. 12 Lutian Road, High-tech Development Zone, Changsha City, Hunan Province, 410205 Applicant before: Hunan soft gold Biotechnology Co.,Ltd. |