Pumpkin germplasm salt tolerance intelligent inspection device

Abstract

The invention discloses a pumpkin germplasm salt tolerance intelligent inspection device which comprises an intelligent trolley, a magnetic strip sensor, a CCD camera, an intelligent holder, a stepping motor controller, a stepping motor, a full-spectrum LED lamp and an intelligent inspection rating module. According to the device, the integrated process of patrolling, stopping, identifying, shooting, transmitting, evaluating and patrolling and collecting of pumpkin seedlings is achieved in the mode of automatic tracking and precise parking of the magnetic strip sensor, the automatic detection degree of the salt stress of the pumpkin seedlings is improved, and precise identification of an image focus area and high-precision automatic focusing are achieved. And the rating module obtains a pumpkin seedling binary mask image through an adaptive segmentation image algorithm, designs a salt stress state index BSD to represent the salt stress degree, and constructs a pumpkin seedling salt stress grade evaluation model, which is suitable for multi-scene accurate salt tolerance rating. The upper computer and the lower computer based on QT cooperate with each other, work is accurate, various functions are integrated, operation is simple, convenient and intelligent, and the automation level and the intelligent level of intelligent agricultural equipment are effectively improved.

Description

Technical Field

[0001] The present invention belongs to the fields of smart agricultural equipment technology, special instruments for crop biological feature identification, crop breeding and selection, and agricultural automation, and specifically relates to an intelligent inspection device for salt tolerance of pumpkin germplasm. Background Art

[0002] Pumpkin is a crop of the genus Cucurbitaceae. It has strong stress resistance, well-developed root system, and wide adaptability. It has been selected as the grafting rootstock for many melon vegetables. However, the accumulation of salt causes pumpkin seedlings to grow slowly, leaves to turn yellow, and even die. It is urgent to screen out salt-resistant pumpkin varieties. Therefore, it is very necessary to detect the salt tolerance of pumpkin seedlings. At present, the methods for judging the degree of salt stress of pumpkin are mainly manual rating and laboratory physical and chemical analysis. However, these methods have high labor costs, poor timeliness, and the results are easily affected by subjective judgment, low accuracy, and poor repeatability. These problems make traditional detection methods unable to meet the needs of large-scale planting and breeding experiments for efficient and accurate data, and more advanced detection technologies are urgently needed. At present, the market needs a more scientific and convenient method to detect the degree of salt stress of pumpkin seedlings, which is convenient for subsequent selection and breeding.

[0003] The technical means of crop salt stress detection research have made certain progress, covering multiple fields and methods. In recent years, the application of hyperspectral imaging technology has provided new possibilities for monitoring plants under salt stress. Wang Qi used spectral information to evaluate the salt tolerance level of okra, providing certain technical support for crop salt stress detection. In addition, image processing technology has also been widely used in the assessment of plant health. Yin Jiandao et al. applied image processing technology to the detection of salt stress response of hydroponic plant roots. Li Yanli et al. used image processing technology to provide a new method for in situ observation of crop roots in salinized areas. With the increase in data volume, modeling methods based on machine learning and deep learning have also been widely used in the study of crop stress detection. Lu Wei's research reviewed the application of deep learning in plant stress detection, showing its potential in salt stress assessment. Zhou Jiayi provided technical support for the establishment of wheat seed germination vitality rating and salt-tolerant wheat breeding by building a PYTORCH deep learning framework.

[0004] Although current research has provided many methods and references for salt stress detection, current research still lacks specialized equipment for pumpkin seedling detection, which makes it difficult to meet the needs of actual agricultural production. In addition, since the leaf shape of pumpkin seedlings is relatively complex after being subjected to salt stress, there is no scientific and accurate indicator for the image characterization of the salt stress state of pumpkin seedlings, and there is no relevant report on the detection of pumpkin seedling salt stress based on the crop phenotyping platform. Summary of the invention

[0005] In order to scientifically and conveniently identify the salt stress degree of pumpkin seedlings, which is conducive to experimenters to better select and breed seeds, the present invention provides an intelligent inspection device for pumpkin germplasm salt tolerance, which realizes the integrated process of pumpkin seedling salt tolerance detection patrol → stop → identify → shoot → transmit → evaluate → patrol and collect, and realizes the automated, rapid and convenient detection of the salt stress degree of pumpkin seedlings.

[0006] In order to solve the above problems, the present invention provides the following technical solutions and proposes an intelligent inspection device for salt tolerance of pumpkin germplasm, which is as follows.

[0007] A pumpkin germplasm salt tolerance intelligent inspection device is used to achieve efficient detection and accurate rating of pumpkin seedling salt tolerance, the inspection device comprises an intelligent car, a magnetic stripe sensor, a CCD camera, an intelligent pan-tilt head, a stepper motor controller, a stepper motor, a full-spectrum LED lamp and an intelligent inspection rating module; wherein the intelligent car works based on the Raspberry Pi and ROS system, and has both remote control and automatic cruise working modes, and as a carrying platform, it can realize free movement in the inspection area; the magnetic stripe sensor is installed on the intelligent car and connected to the Raspberry Pi through a serial port to control automatic cruise and precise parking; the CCD camera is fixed on the intelligent pan-tilt head, and the intelligent pan-tilt head is installed on the screw slide carried by the intelligent car through a telescopic rod that can move up and down, and the intelligent pan-tilt head is used to stabilize the CCD camera and keep it vertical Shooting downward, the CCD camera is used to capture the identity QR code information and the image of the pumpkin seedling; the stepper motor controller is fixed to the lower position of the screw slide carried by the smart car by screws, and is used to control the lifting of the stepper motor; the stepper motor is fixed to the uppermost end of the screw slide carried by the smart car, and a pulse signal is sent through the serial port to control the lifting of the smart pan / tilt and the CCD camera to obtain the best field of view; the full-spectrum LED light is fixed to the middle position of the screw slide carried by the smart car by a connecting rod, adopts a full-field rotatable design, is connected to the Raspberry Pi through a relay, and the LED light can be controlled by the host computer to ensure uniform lighting when shooting the pumpkin seedling image; the intelligent inspection and rating module is mounted on the smart car, and the salt stress level assessment of the pumpkin seedlings is completed through image processing and calculation.

[0008] Preferably, the Raspberry Pi of the smart car is programmed through PYCHARM, and the SSH protocol is used to remotely control the lifting of the stepper motor, the movement of the car, the CCD camera and the full-spectrum LED light; the host computer remotely controls the Raspberry Pi to achieve comprehensive control of the entire inspection device.

[0009] Preferably, the magnetic stripe sensor obtains the magnetic stripe position information to realize the automatic patrol of the smart car on the laid magnetic stripe track, specifically in the following manner: Step S1: During the working process, an industrial magnetic stripe track is laid on the ground, and the magnetic stripe sensor feeds back the position information of the magnetic stripe in real time; Step S2, attaching a magnetic strip perpendicular to the patrol track next to each pot of pumpkin seedlings; Step S3, the smart car establishes a connection with the magnetic stripe sensor through serial communication, and then makes a choice of the movement of the smart car to ensure that the smart car patrols along the magnetic stripe track; Step S4, when the smart car drives to the side of the pumpkin seedling and sticks the magnetic stripe horizontally, the switch points on the left and right sides of the magnetic stripe sensor operate, and the car obtains the operating status feedback information of the 16 switch points, executes the stop command, and achieves precise parking.

[0010] Preferably, the CCD camera obtains appropriate height parameters in the following manner: The minimum working height of the CCD camera is calculated using the following formula: FOV min Indicates the minimum field of view of the CCD camera, h min represents the minimum working height distance of the CCD camera, and θ represents the field of view angle of the CCD camera; The maximum working height of the CCD camera is calculated using the following formula: Where P fix Represents the pixel of CCD camera, P min Indicates the minimum pixel requirement, S lea Indicates the leaf area, FOV max Indicates the maximum field of view of the CCD camera required for shooting, h max represents the maximum working height distance of the CCD camera, and θ represents the field of view angle of the CCD camera; After obtaining the minimum working height and maximum working height of the CCD camera, the parameters are fed back to the host computer, and the height of the CCD camera is automatically adjusted by the stepper motor.

[0011] Preferably, when the CCD camera takes the image of the pumpkin seedling, an automatic focusing algorithm based on the Hilbert space angle is used to control the stepper motor to lift the CCD for focusing, specifically in the following manner: Step S1, obtaining the first pumpkin seedling image at the highest point of the lifting platform; Step S2, the stepper motor drives the lifting platform to move downward by 1 mm, obtains the second image, and calculates the Hilbert space angle θ1 by the following formula; Where θ1 is the Hilbert space angle of the two images. The larger the θ1 value, the smaller the similarity between the two images. x, y are the pixel coordinates of the image, R is the image pixel range, and F(x, y) and G(x, y) are the intensities of the corresponding pixels in the two images. Step S3: The stepper motor continues to drive the lifting platform to move i 1mm in the same direction to obtain the i+1th image. The Hilbert space angle between the first image and the i+1th image is calculated to obtain the value θ i ; Step S4, when θ i When it reaches the peak, the image of the pumpkin seedling is in the clearest focus and the CCD camera achieves focus.

[0012] Preferably, the intelligent inspection rating module is based on QT design host computer interface to carry out integrated control of intelligent inspection device, integrate image color correction, background segmentation, salt stress feature extraction, calculate salt stress state index BSD inversion pumpkin seedling salt stress level steps, complete pumpkin seedling salt stress level assessment, specifically adopt the following method, Step S1, using a 24-color standard color card to correct the color changes caused by different lighting environments, regressing the chromaticity values ​​extracted from the photographed 24-color standard color card with the chromaticity values ​​of the real colorimetric card to obtain an equation corresponding to each band, substituting the equation into each picture for calculation to correct the pumpkin seedling image; Step S2, pumpkin seedling image background and leaf segmentation is performed using the following formula: Among them, R, G, and B correspond to the grayscale values ​​of the R, G, and B channels in the pumpkin seedling image, respectively. According to the double peaks of the γ value corresponding to the background and leaves of the pumpkin seedling image, the critical position of the two peaks is selected as the threshold by using the adaptive segmentation algorithm to perform background segmentation and generate a pumpkin seedling binary mask image; Step S3, the salt stress state index BSD is calculated using the following formula to obtain the salt stress characteristics of pumpkin seedlings: Where P(R MAX )、P(G MAX )、P(B MAX ) correspond to the most probable values ​​of the RGB three-channel intensity respectively; Step S4, setting the salt damage index SI as the dependent variable and the salt stress state index BSD as the independent variable to establish a pumpkin seedling salt stress grade assessment model, and using the following formula to calculate the salt damage index SI to rate the salt tolerance of the pumpkin seedlings: SI=-2.463*BSD+1.151 (BSD<0.35) SI=0.276 (BSD≥0.35) Wherein, BSD is the salt stress state index calculated in step S3, and SI is the salt damage index, which is used to characterize the salt damage level. When SI is less than 35%, the salt tolerance is strong; when 35%≤SI<65%, the salt tolerance is medium; and when SI≥65%, the salt tolerance is weak.

[0013] Compared with the prior art, the present invention has significant positive technical effects, which are specifically manifested in the following aspects.

[0014] (1) In the prior art, the identification of salt stress in pumpkin seedlings mainly relies on manual rating and laboratory physical and chemical analysis, which is time-consuming, labor-intensive, inefficient, and costly, making it difficult to meet the needs of large-scale planting and breeding. The present invention integrates an intelligent vehicle, a magnetic stripe sensor, a CCD camera, an intelligent pan / tilt, a stepper motor controller, a stepper motor, a full-spectrum LED light, and an intelligent patrol rating module to realize an intelligent integrated process of patrol → stop → identify → shoot → transmit → evaluate → patrol, which greatly improves the automation and intelligence level of crop trait detection devices. The intelligent vehicle uses a magnetic stripe sensor to obtain the magnetic stripe position information to realize automatic patrol ( patrol ), the running status information of the horizontal magnetic strip feedback switch point is detected during driving to achieve precise parking ( stop ), identify the pumpkin seedlings’ identity QR code and record the sample number ( knowledge ), turn on the full-spectrum LED light and use the CCD camera to capture RGB images ( shoot ), and pass the pumpkin seedling sample number and RGB image to the host computer interface for display ( pass ), the pumpkin seedling salt tolerance evaluation module automatically rates salt damage and displays ( Comments ), and then continue to patrol automatically. The present invention effectively improves the detection efficiency and accuracy, and realizes a fast and accurate evaluation of the salt stress level of pumpkin seedlings.

[0015] (2) Most of the existing image autofocus technologies rely on traditional gradient or spectrum analysis, which has the problems of complex algorithms and low efficiency. This application designs an autofocus method based on Hilbert space angles, and effectively solves the problems of low focusing efficiency and poor robustness in the existing technology by calculating the angle information of the image in the Hilbert space, and realizes accurate identification of the focus area of ​​the image, achieving the effect of high-precision autofocus.

[0016] (3) In the prior art, the rating of pumpkin seedlings mainly relies on manual evaluation, usually judged by visual inspection or simple quantitative indicators, which is easily affected by human subjective factors, resulting in poor accuracy and reliability of the rating results. This application designs a pumpkin seedling salt stress level assessment module based on the salt stress state index BSD, specifically integrating image color correction, background segmentation, salt stress feature extraction, and calculation of the salt stress state index BSD to invert the pumpkin seedling salt stress level steps, effectively quantifying the pumpkin seedling salt stress process, improving the accuracy and consistency of the rating, and achieving accurate and objective identification of the degree of salt stress of pumpkin seedlings. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a schematic diagram of the device structure of the present invention.

[0018] Figure 2 It is a schematic diagram of the collection work of the present invention.

[0019] Figure 3 It is a schematic diagram of the automatic focusing method of the present invention.

[0020] Figure 4 This is a graph showing the accuracy verification results of the pumpkin seedling salt stress rating model used in the present invention.

[0021] Figure 5 This is a graph showing the performance test results of the pumpkin seedling salt stress rating model used in the present invention.

[0022] in, Figure 1 Among them, 1-intelligent pan / tilt and CCD camera, 2-full spectrum LED light, 3-lifting platform, 4-magnetic strip sensor, 5-intelligent car. DETAILED DESCRIPTION

[0023] The present invention is further described below in conjunction with the accompanying drawings and implementation examples.

[0024] A pumpkin germplasm salt tolerance intelligent inspection device, such as Figure 1As shown, it is used to realize efficient detection and accurate rating of salt tolerance of pumpkin seedlings. The inspection device includes an intelligent car, a magnetic stripe sensor, a CCD camera, an intelligent pan-tilt, a stepper motor controller, a stepper motor, a full-spectrum LED lamp, and an intelligent inspection rating module; wherein the intelligent car works based on the Raspberry Pi and ROS system, and has both remote control and automatic cruise working modes. As a carrying platform, it can realize free movement in the inspection area; the magnetic stripe sensor is installed on the intelligent car and connected to the Raspberry Pi through a serial port to control automatic cruise and accurate parking; the CCD camera is fixed on the intelligent pan-tilt, and the intelligent pan-tilt is installed on the screw slide carried by the intelligent car through a telescopic rod that can move up and down. The intelligent pan-tilt is used to stabilize the CCD camera and keep it shooting vertically downward. The CD camera is used to capture identity QR code information and images of pumpkin seedlings; the stepper motor controller is fixed to the lower position of the screw slide carried by the smart car by screws, and is used to control the lifting of the stepper motor; the stepper motor is fixed to the uppermost end of the screw slide carried by the smart car, and sends pulse signals through the serial port to control the lifting of the smart pan / tilt and CCD camera to obtain the best field of view; the full-spectrum LED lamp is fixed to the middle position of the screw slide carried by the smart car by a connecting rod, adopts a full-field rotatable design, is connected to the Raspberry Pi by a relay, and the LED lamp can be controlled by the host computer to ensure uniform lighting when shooting images of pumpkin seedlings; the intelligent inspection and rating module is mounted on the smart car, and completes the salt stress level assessment of pumpkin seedlings through image processing and calculation.

[0025] Preferably, the Raspberry Pi of the smart car is programmed through PYCHARM, and the SSH protocol is used to remotely control the lifting of the stepper motor, the movement of the car, the CCD camera and the full-spectrum LED light; the host computer remotely controls the Raspberry Pi to achieve comprehensive control of the entire inspection device.

[0026] Preferably, the magnetic stripe sensor obtains the magnetic stripe position information to realize the automatic patrol of the smart car on the laid magnetic stripe track, such as Figure 2 As shown, the specific method is as follows: Step S1: During the working process, an industrial magnetic stripe track is laid on the ground, and the magnetic stripe sensor feeds back the position information of the magnetic stripe in real time; Step S2, attaching a magnetic strip perpendicular to the patrol track next to each pot of pumpkin seedlings; Step S3, the smart car establishes a connection with the magnetic stripe sensor through serial communication, and then makes a choice of the movement of the smart car to ensure that the smart car patrols along the magnetic stripe track; Step S4, when the smart car drives to the side of the pumpkin seedling and sticks the magnetic stripe horizontally, the switch points on the left and right sides of the magnetic stripe sensor operate, and the car obtains the operating status feedback information of the 16 switch points, executes the stop command, and achieves precise parking.

[0027] Preferably, the CCD camera obtains appropriate height parameters in the following manner: The minimum working height of the CCD camera is calculated using the following formula: FOV min Indicates the minimum field of view of the CCD camera, h min represents the minimum working height distance of the CCD camera, and θ represents the field of view angle of the CCD camera; The maximum working height of the CCD camera is calculated using the following formula: Where P fix Represents the pixel of CCD camera, P min Indicates the minimum pixel requirement, S lea Indicates the leaf area, FOV max Indicates the maximum field of view of the CCD camera required for shooting, h max represents the maximum working height distance of the CCD camera, and θ represents the field of view angle of the CCD camera; After obtaining the minimum working height and maximum working height of the CCD camera, the parameters are fed back to the host computer, and the height of the CCD camera is automatically adjusted by the stepper motor.

[0028] Preferably, when the CCD camera takes the image of the pumpkin seedling, an automatic focusing algorithm based on the Hilbert space angle is used to control the stepper motor to lift the CCD for focusing, specifically in the following manner: Step S1, obtaining the first pumpkin seedling image at the highest point of the lifting platform; Step S2, the stepper motor drives the lifting platform to move downward by 1 mm, obtains the second image, and calculates the Hilbert space angle θ1 by the following formula; Where θ1 is the Hilbert space angle of the two images. The larger the θ1 value, the smaller the similarity between the two images. x, y are the pixel coordinates of the image, R is the image pixel range, and F(x, y) and G(x, y) are the intensities of the corresponding pixels in the two images. Step S3: The stepper motor continues to drive the lifting platform to move i 1mm in the same direction to obtain the i+1th image. The Hilbert space angle between the first image and the i+1th image is calculated to obtain the value θ i ; Step S4, when θ i When it reaches the peak, the image of the pumpkin seedling is in the clearest focus and the CCD camera achieves focus.

[0029] Preferably, the intelligent inspection rating module is based on QT design host computer interface to carry out integrated control of intelligent inspection device, integrate image color correction, background segmentation, salt stress feature extraction, calculate salt stress state index BSD inversion pumpkin seedling salt stress level steps, complete pumpkin seedling salt stress level assessment, specifically adopt the following method, Step S1, using a 24-color standard color card to correct the color changes caused by different lighting environments, regressing the chromaticity values ​​extracted from the photographed 24-color standard color card with the chromaticity values ​​of the real colorimetric card to obtain an equation corresponding to each band, substituting the equation into each picture for calculation to correct the pumpkin seedling image; Step S2, pumpkin seedling image background and leaf segmentation is performed using the following formula: Among them, R, G, and B correspond to the grayscale values ​​of the R, G, and B channels in the pumpkin seedling image, respectively. According to the double peaks of the γ value corresponding to the background and leaves of the pumpkin seedling image, the critical position of the two peaks is selected as the threshold by using the adaptive segmentation algorithm to perform background segmentation and generate a pumpkin seedling binary mask image; Step S3, the salt stress state index BSD is calculated using the following formula to obtain the salt stress characteristics of pumpkin seedlings: Where P(R MAX )、P(G MAX )、P(B MAX ) correspond to the most probable values ​​of the RGB three-channel intensity respectively; Step S4, setting the salt damage index SI as the dependent variable and the salt stress state index BSD as the independent variable to establish a pumpkin seedling salt stress grade assessment model, and using the following formula to calculate the salt damage index SI to rate the salt tolerance of the pumpkin seedlings: SI=-2.463*BSD+1.151 (BSD<0.35) SI=0.276 (BSD≥0.35) Wherein, BSD is the salt stress state index calculated in step S3, and SI is the salt damage index, which is used to characterize the salt damage level. When SI is less than 35%, the salt tolerance is strong; when 35%≤SI<65%, the salt tolerance is medium; and when SI≥65%, the salt tolerance is weak.

[0030] In order to better explain the technical solution of the present invention, taking pumpkin seedlings as specific application objects, the present invention obtained the following results.

[0031] (1) The intelligent inspection device for pumpkin germplasm salt tolerance realizes the integrated process of inspection → stop → identify → photograph → transmit → evaluate → inspect and collect pumpkin seedlings, thus improving the level of automated detection of pumpkin seedling salt stress.

[0032] (2) Based on the scheme proposed by the present invention, salt stress identification was performed on the true leaves of pumpkin seedlings, such as Figure 4 As shown, the salt stress rating model R of pumpkin seedlings 2 The prediction accuracy is 84.79%, which can effectively detect the salt stress degree of pumpkin seedlings.

[0033] (3) 49 pumpkin varieties were selected, and the corresponding BSD mean values ​​were calculated. The pumpkin seedling salt stress rating model constructed using the BSD index was used to rate the salt tolerance of each pumpkin seedling material. The salt tolerance grades of each pumpkin variety were compared with those obtained using the three-level salt tolerance classification standard (strong salt tolerance SI < 35%, medium 35% ≤ SI < 65%, weak SI ≥ 65%). The rating accuracy calculated using the formula shown below was 89.79%. The calculation results are as follows: Figure 5 As shown, it can be found that the present invention can rate the salt stress of pumpkin seedlings more accurately.

[0034] The specific examples described in the application are merely examples of the spirit of the present invention. Those skilled in the art may make various modifications or additions to the specific examples described in the present invention, or replace them in a similar manner, but they will not deviate from the spirit of the present invention or exceed the scope defined by the attached claims.

Claims

1. An intelligent inspection device for salt tolerance of pumpkin germplasm, used to achieve efficient detection and accurate rating of salt tolerance of pumpkin seedlings, characterized in that: The inspection device includes an intelligent car, a magnetic stripe sensor, a CCD camera, an intelligent pan-tilt head, a stepper motor controller, a stepper motor, a full-spectrum LED light and an intelligent inspection rating module; wherein, the intelligent car works based on the Raspberry Pi and ROS system, and has both remote control and automatic cruise working modes. As a carrying platform, it can realize free movement in the inspection area; the magnetic stripe sensor is installed on the intelligent car and connected to the Raspberry Pi through a serial port to control automatic cruise and precise parking; the CCD camera is fixed on the intelligent pan-tilt head, and the intelligent pan-tilt head is installed on the screw slide carried by the intelligent car through a telescopic rod that can move up and down. The intelligent pan-tilt head is used to stabilize the CCD camera and keep it vertically downward for shooting. The CCD camera is used to shoot the identity QR code The stepper motor controller is fixed to the lower position of the screw slide carried by the smart car by screws to control the lifting of the stepper motor; the stepper motor is fixed to the uppermost end of the screw slide carried by the smart car, and a pulse signal is sent through the serial port to control the lifting of the smart pan / tilt and CCD camera to obtain the best field of view; the full-spectrum LED lamp is fixed to the middle position of the screw slide carried by the smart car by a connecting rod, adopts a full-field rotatable design, is connected to the Raspberry Pi by a relay, and the LED lamp can be controlled by the host computer to ensure uniform lighting when taking images of pumpkin seedlings; the intelligent inspection and rating module is mounted on the smart car, and the salt stress level assessment of pumpkin seedlings is completed through image processing and calculation.

2. The pumpkin germplasm salt tolerance intelligent inspection device according to claim 1, characterized in that: The Raspberry Pi of the smart car is programmed through PYCHARM, and uses the SSH protocol to remotely control the lifting of the stepper motor, the movement of the car, the CCD camera and the full-spectrum LED light; the host computer remotely controls the Raspberry Pi to achieve comprehensive control of the entire inspection device.

3. The pumpkin germplasm salt tolerance intelligent inspection device according to claim 1, characterized in that: The magnetic stripe sensor obtains the magnetic stripe position information to realize the automatic patrol of the intelligent car on the laid magnetic stripe track, specifically in the following manner: Step S1: During the working process, an industrial magnetic stripe track is laid on the ground, and the magnetic stripe sensor feeds back the position information of the magnetic stripe in real time; Step S2, attaching a magnetic strip perpendicular to the patrol track next to each pot of pumpkin seedlings; Step S3, the smart car establishes a connection with the magnetic stripe sensor through serial communication, and then makes a choice of the movement of the smart car to ensure that the smart car patrols along the magnetic stripe track; Step S4, when the smart car drives to the side of the pumpkin seedlings and sticks the magnetic stripe horizontally, the switch points on the left and right sides of the magnetic stripe sensor operate, and the car obtains the operating status feedback information of the 16 switch points, executes the stop command, and achieves precise parking.

4. The pumpkin germplasm salt tolerance intelligent inspection device according to claim 1, characterized in that: The CCD camera obtains the appropriate height parameters in the following manner: The minimum working height of the CCD camera is calculated using the following formula: FOV min Indicates the minimum field of view of the CCD camera, h min represents the minimum working height distance of the CCD camera, and θ represents the field of view angle of the CCD camera; The maximum working height of the CCD camera is calculated using the following formula: Where P fix Represents the pixel of CCD camera, P min Indicates the minimum pixel requirement, S lea Indicates the leaf area, FOV max Indicates the maximum field of view of the CCD camera required for shooting, h max Indicates the maximum working height distance of the CCD camera; After obtaining the minimum working height and maximum working height of the CCD camera, the parameters are fed back to the host computer, and the height of the CCD camera is automatically adjusted by the stepper motor.

5. The pumpkin germplasm salt tolerance intelligent inspection device according to claim 1, characterized in that: When the CCD camera takes pictures of pumpkin seedlings, it uses an automatic focusing algorithm based on the Hilbert spatial angle to control the stepper motor to lift the CCD for focusing. The specific method is as follows: Step S1, obtaining the first pumpkin seedling image at the highest point of the lifting platform; Step S2: The stepper motor drives the lifting platform to move downward by 1 mm to obtain the second image. The Hilbert space angle θ is calculated by the following formula: 1， θ 1 = a r c c o s ∑ x , y ∈ R F x y G x y ∑ x , y ∈ R F x y 2 1 2 ∑ x , y ∈ R G x y 2 1 2 Where θ1 is the Hilbert space angle of the two images. The larger the θ1 value, the smaller the similarity between the two images. x, y are the pixel coordinates of the image, R is the pixel range of the image, and F(x, y) and G(x, y) are the intensities of the corresponding pixels in the two images. Step S3: The stepper motor continues to drive the lifting platform to move i 1mm in the same direction to obtain the i+1th image. The Hilbert space angle between the first image and the i+1th image is calculated to obtain the value θ i ; Step S4, when θ i When it reaches the peak, the image of the pumpkin seedling is in the clearest focus and the CCD camera achieves focus.

6. The pumpkin germplasm salt tolerance intelligent inspection device according to claim 1, characterized in that: The intelligent inspection rating module is based on QT design of the host computer interface to carry out integrated control of the intelligent inspection device, integrate image color correction, background segmentation, salt stress feature extraction, calculate the salt stress state index BSD inversion pumpkin seedling salt stress level steps, and complete the pumpkin seedling salt stress level assessment, specifically using the following method: Step S1, using a 24-color standard color card to correct the color changes caused by different lighting environments, regressing the chromaticity values ​​extracted from the photographed 24-color standard color card with the chromaticity values ​​of the real colorimetric card to obtain an equation corresponding to each band, substituting the equation into each picture for calculation to correct the pumpkin seedling image; Step S2, pumpkin seedling image background and leaf segmentation is performed using the following formula: Among them, R, G, and B correspond to the grayscale values ​​of the R, G, and B channels in the pumpkin seedling image, respectively. According to the double peaks of the γ value corresponding to the background and leaves of the pumpkin seedling image, the critical position of the two peaks is selected as the threshold by using the adaptive segmentation algorithm to perform background segmentation and generate a pumpkin seedling binary mask image; Step S3, the salt stress state index BSD is calculated using the following formula to obtain the salt stress characteristics of pumpkin seedlings: Where P(R MAX )、P(G MAX )、P(B MAX ) correspond to the most probable values ​​of the RGB three-channel intensity respectively; Step S4, setting the salt damage index SI as the dependent variable and the salt stress state index BSD as the independent variable to establish a pumpkin seedling salt stress grade assessment model, and using the following relationship to calculate the salt damage index SI to rate the salt tolerance of the pumpkin seedlings: SI=-2.463*BSD+1.151 (BSD<0.35) SI=0.276 (BSD≥0.35) Wherein, BSD is the salt stress state index calculated in step S3, and SI is the salt damage index, which is used to characterize the salt damage level. When SI is less than 35%, the salt tolerance is strong; when 35%≤SI<65%, the salt tolerance is medium; and when SI≥65%, the salt tolerance is weak.

Images