Plant image acquisition device, plant quality monitoring system and method

Abstract

The application provides a plant image acquisition device, a plant quality monitoring system and a method. The plant quality monitoring system comprises a plant image acquisition device, a control module configured to control the plant image acquisition device to capture a color image and a grayscale image under a preset wave band of a plant in a target culture plate at the same time, and a processing module configured to determine a plant index score, a plant weight and a plant phenotype proportion of the plant in the target culture plate according to the images captured by the plant image acquisition device, and determine a quality level of the plant in the target culture plate according to at least one of the plant index score, the plant weight and the target plant phenotype proportion. The problem of low plant quality evaluation accuracy is solved.

Description

Technical Field

[0001] This invention relates to the field of agricultural information technology, specifically providing a plant image acquisition device, a plant quality monitoring system, and a method. Background Technology

[0002] In industrialized plant production, plant quality monitoring and assessment are closely related to product quality, and digitalized industrial monitoring of plant quality improves work efficiency. However, the industry currently faces major challenges, including low accuracy in plant quality assessment and difficulties in data traceability.

[0003] The main reasons for the low accuracy of plant quality monitoring are as follows: (1) Traditional forage quality assessment relies on manual visual inspection, which can only make rough judgments with large errors; (2) External stray light interference and uneven lighting lead to large fluctuations in the quality of the collected images, affecting the accuracy of subsequent data analysis; (3) Existing segmentation models have weak ability to distinguish plant phenotypic regions, especially for bald patches and moldy areas with blurred boundaries, resulting in high segmentation errors and inaccurate quality assessment; (4) The assessment dimension is singular and cannot fully reflect the quality of forage.

[0004] In addition, the difficulty in data tracing is mainly due to the chaotic management of images taken from multiple angles, which makes it difficult to find data during subsequent data analysis. Summary of the Invention

[0005] The technical problem to be solved by this invention is to overcome the shortcomings of the prior art. The purpose is to provide a plant image acquisition device that, by setting up an acquisition box and a transmission unit, avoids the manual handling of the culture tray from the conveyor belt to the acquisition box, eliminates the need for intermediate transfer links, greatly improves processing efficiency, and enhances industrialization. Furthermore, this invention provides a plant quality monitoring system and method that, through at least one of phenotypic segmentation, weight prediction, and plant index, determines the quality level of the plants in the target culture tray, constructs a more complete quality evaluation standard, and improves the accuracy of plant quality assessment.

[0006] The first objective of this invention is to provide a plant image acquisition device, comprising: an acquisition box, wherein an image acquisition unit is disposed within the acquisition box; a transmission unit for transmitting a plant culture tray; the acquisition box having a closable opening for docking with the transmission unit, so that the image acquisition unit acquires images of the culture tray transported to the acquisition box.

[0007] Furthermore, the opening includes an inlet and an outlet; the conveying unit delivers the culture tray into the acquisition box through the inlet and delivers the culture tray out of the acquisition box through the outlet; the image acquisition unit is located on the upper part of the acquisition box; preferably, the number of the image acquisition units is at least two, which are used to acquire color images of plants and grayscale images under preset bands, respectively.

[0008] Furthermore, the acquisition box includes a box frame and a base plate; the box frame is located above the transmission unit; the base plate surrounds at least part of the box frame and forms the acquisition box above the transmission unit.

[0009] Furthermore, it also includes a culture rack, wherein at least a portion of the culture rack is provided with a conveying unit; the culture rack includes columns that intersect the conveying unit in the vertical direction; at least a portion of the columns are used to form the frame of the collection box.

[0010] Furthermore, the image acquisition unit is a near-infrared camera, and at least some of the image acquisition units are configured with narrow-band filters whose center band is the preset band; preferably, the acquisition box also includes a light source assembly for providing a full-spectrum light source and an infrared light source; more preferably, the light source assembly is located on the upper part of the acquisition box and distributed around the near-infrared camera assembly.

[0011] The second objective of this invention is to provide a plant image acquisition system, including the aforementioned plant image acquisition device, and further comprising: a control module for controlling the plant image acquisition device to simultaneously capture color images and grayscale images of plants in a target culture tray under a preset wavelength; and a processing module for determining, based on the images captured by the plant image acquisition device, the plant index score, plant weight, and plant phenotypic percentage of the plants in the target culture tray; and determining the quality grade of the plants in the target culture tray based on at least one of the plant index score, plant weight, and target plant phenotypic percentage.

[0012] Furthermore, the processing module is also used to extract image features from the image and calculate a plant index score and / or plant weight based on the extracted image features; preferably, the plant weight is calculated based on image features of a preset ratio and a specific preset band; preferably, the plant index score is calculated based on image features of a color image and / or a grayscale image under a preset band using a preset formula; more preferably, the plant index includes at least two of the following: NDVI, GNDVI, NDGI, CIG, GLI, MCARI, MTVI2.

[0013] Furthermore, the processing module is also used to segment the image based on the trained image segmentation model, wherein the segmented image includes at least two phenotypic feature regions; and to determine the phenotypic proportion of the target plant based on the area of ​​the phenotypic feature regions; preferably, the phenotypic feature regions include healthy regions, mild alopecia areata regions, severe alopecia areata regions, and fungal infection regions.

[0014] Furthermore, it also includes a display interaction module for receiving user input; the processing module is also used to determine the target culture tray based on the user input; preferably, the display interaction module is also used to display the quality grade of the plant in the target culture tray and / or the segmented image; more preferably, it also includes a storage module for storing the quality grade of the plant in the target culture tray and / or the image; the processing module is also used to determine the plant information to be photographed and the camera information for photographing the plant based on the user input; and to name the captured image based on the plant information and the camera information.

[0015] The third objective of this invention is to provide a plant quality monitoring method applied to the aforementioned plant quality monitoring system, comprising: simultaneously capturing a color image and a grayscale image of a plant in a target culture tray at a preset wavelength; determining, based on the captured images, a plant index, plant weight, and plant phenotypic percentage corresponding to the plant in the target culture tray; and determining, based on at least one of the plant index, plant weight, and target plant phenotypic percentage, the quality grade of the plant in the target culture tray.

[0016] In summary, the present invention provides a plant image acquisition device, a plant quality monitoring system, and a method, which have the following advantages compared with the prior art: The plant image acquisition device, composed of a collection box and a conveyor unit, eliminates the need for manual handling of culture trays from the conveyor belt to the collection box, removing intermediate transfer links and significantly improving processing efficiency. It upgrades the single acquisition function into a core component of industrial production, adapting to the needs of large-scale production and supporting production line linkage and intelligent decision-making. The use of the culture rack column support as the framework of the collection box also saves costs. The closable opening on the collection box reduces external light interference, providing a high-quality data source for spectral analysis and phenotypic segmentation. Phenotypic segmentation, weight prediction, and plant indices provide data support for a comprehensive quantitative assessment of plant quality. Based on at least one of the following: plant index score, plant weight, and the proportion of the target plant phenotype, the quality level of the plant in the target culture tray is determined, constructing a more comprehensive quality assessment standard, improving the coverage of assessment dimensions, thereby increasing the accuracy of results and overcoming the problem of misjudgment that can easily occur when relying on a single indicator. Attached Figure Description

[0017] The accompanying drawings, as part of this invention, are provided to further illustrate the invention. The illustrative embodiments and descriptions of the invention are used to explain the invention, but do not constitute an undue limitation thereof. Clearly, the drawings described below are merely some embodiments, and those skilled in the art can obtain other drawings based on these drawings without any creative effort.

[0018] In the attached diagram: Figure 1 This is a side view of a plant image acquisition device provided by the present invention; Figure 2 This is a front view of another plant image acquisition device provided by the present invention; Figure 3 This is a schematic diagram of the structure of a data collection box provided by the present invention; Figure 4 This is a schematic diagram of the structure of a plant image acquisition device provided by the present invention; Figure 5 This is a schematic diagram of the component distribution inside a data acquisition box provided by the present invention; Figure 6 This is a flowchart of a plant quality monitoring method provided by the present invention.

[0019] In the figure, 1-acquisition box; 11-opening; 12-box frame; 13-substrate; 2-transfer unit; 3-image acquisition unit; 4-culture tray; 5-column; 6-light source assembly.

[0020] It should be noted that these accompanying drawings and textual descriptions are not intended to limit the scope of the invention in any way, but rather to illustrate the concept of the invention to those skilled in the art by referring to specific embodiments. Detailed Implementation

[0021] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0022] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0023] like Figure 1-2As shown, the present invention provides a plant image acquisition device, comprising: an acquisition box 1, wherein an image acquisition unit 3 is disposed inside the acquisition box 1; a transmission unit 2 for transmitting a plant culture tray 4; the acquisition box 1 is provided with a closable opening 11, the opening 11 for docking with the transmission unit 2, so that the image acquisition unit 3 acquires images of the culture tray 4 transported to the acquisition box 1.

[0024] Specifically, the acquisition box can cut off external light sources, thus reducing interference from external light during the shooting process.

[0025] Because external light includes multiple wavelengths and has varied incident angles, it can easily create strong reflections or shadows on the plant surface, leading to localized grayscale distortion in the captured images and failing to reflect the plant's true optical characteristics. This resulted in the misclassification of healthy areas as diseased areas during subsequent quality assessments.

[0026] It should be noted that the collection box can be a completely enclosed box structure or a cover structure formed by the surrounding walls and the top wall.

[0027] Specifically, the image acquisition unit can be a camera with its lens pointing downwards to photograph the plants in the cultivation tray.

[0028] In one embodiment, the inner wall of the acquisition box is coated with a black coating to block external light and weaken diffused light. In this embodiment, the image acquisition unit also includes a light source to provide illumination for image acquisition.

[0029] In another embodiment, the inner wall of the collection box is made of frosted material. The uneven texture structure changes the light reflection path, preventing light from reflecting off plants or camera lenses and causing local glare.

[0030] Specifically, during the image acquisition process, the opening on the acquisition box is closed to form a closed state; before or after image acquisition, the opening on the acquisition box is opened to allow the culture tray to enter and exit the acquisition box through the opening.

[0031] Specifically, the opening is usually located at the bottom of the collection box.

[0032] Specifically, the opening can open or close automatically, or it can be opened or closed manually.

[0033] Figure 1 The image shown is a side view of the plant image acquisition device. Figure 2 The image shown is a front view of the plant image acquisition device. As shown, the opening 11 of the acquisition box 1 is large enough for the culture tray 4 to enter.

[0034] In one embodiment, the opening diameter of the opening 11 can be adjusted so that culture trays of different widths and plants of different heights can enter and exit the collection box 1.

[0035] Specifically, the conveying unit can be a belt drive structure or a roller / drum drive structure, such as a conveyor belt or a conveyor roller.

[0036] In addition, the transmission unit can be powered by external force or driven by inertia. If the transmission unit is powered by external force, the plant image acquisition device also includes a drive motor to drive the transmission unit, on which the culture tray to be transmitted is placed.

[0037] In addition, the transmission channel of the transmission unit can be a closed loop channel.

[0038] Furthermore, the opening includes an inlet and an outlet; the conveying unit delivers the culture tray into the acquisition box through the inlet and delivers the culture tray out of the acquisition box through the outlet; the image acquisition unit is located on the upper part of the acquisition box; preferably, the number of the image acquisition units is at least two, which are used to acquire color images of plants and grayscale images under preset bands, respectively.

[0039] like Figure 2 As shown, the arrows indicate the conveying direction of the conveying unit, which passes through the inlet and outlet of the collection box along the conveying direction.

[0040] Figure 2 The end with the arrow indicates the upstream of the conveyor unit, and the other end indicates the downstream of the conveyor unit. Accordingly, the inlet and outlet of the collection box can be identified based on the upstream and downstream of the conveyor unit.

[0041] The openings 11 are provided on the two side walls of the collection box 1, serving as the inlet and outlet for the culture tray to enter and exit the collection box. The two openings 11 can be located on two opposite or adjacent side walls of the collection box 1.

[0042] Specifically, the image acquisition unit can be installed on the top wall of the acquisition box, or at the upper end of the side wall of the acquisition box. This ensures that the imaging area covers the culture dish, guaranteeing the integrity of the image.

[0043] In this embodiment, if there are two image acquisition units, one is used to acquire a color image of the culture dish, and the other is used to acquire a grayscale image under a preset wavelength. If there are multiple image acquisition units, one is used to acquire a color image of the culture dish, and the remaining image acquisition units are used to acquire grayscale images under preset wavelengths.

[0044] The color image is used for plant phenotypic detection and region segmentation, while the grayscale image is used to calculate at least one plant index and plant weight, thereby evaluating plant instructions from multiple perspectives and improving the accuracy of instruction evaluation.

[0045] Furthermore, the acquisition box includes a box frame and a base plate; the box frame is located above the transmission unit; the base plate surrounds at least part of the box frame and forms the acquisition box above the transmission unit.

[0046] like Figure 3 As shown, the box frame in this embodiment consists of horizontal beams and vertical beams. In other embodiments, the box frame may only include vertical beams arranged at preset intervals.

[0047] Based on the above, the substrate can be enclosed at the six openings of the box frame to form a fully enclosed box structure; the substrate can also be enclosed around the perimeter and top of the box frame to form a cover structure.

[0048] Specifically, the size of the acquisition box is set according to the size of the culture tray, reserving enough space to set up the image acquisition unit. For example, if the size of the culture tray is 80cm×40cm, then the size of the acquisition box can be 120cm×60cm×80cm.

[0049] In one embodiment, the substrate is made of acrylic. As described above, the side of the acrylic substrate facing the collection box is made of a frosted material and coated with a black coating.

[0050] Furthermore, it also includes a culture rack, wherein at least a portion of the culture rack is provided with a conveying unit; the culture rack includes columns that intersect the conveying unit in the vertical direction; at least a portion of the columns are used to form the frame of the collection box.

[0051] like Figure 4 As shown in the figure, the entire culture rack is not depicted, but the uprights 5 of the culture rack are shown. The uprights 5 intersect the transmission unit 5 in the vertical direction. Furthermore, the arrangement direction of the uprights 5 is perpendicular or parallel to the transmission direction of the transmission unit 2.

[0052] In one embodiment, the culture rack further includes a support plate, which is fixedly connected to the column, and the conveying unit is disposed above the support plate.

[0053] According to the above embodiments, if the acquisition box is a cover structure, the lower end of the side wall of the acquisition box is lower than or level with the support plate in the horizontal direction, and the width of the support plate is greater than or equal to the width of the side wall of the acquisition box, so that the acquisition box and the support plate together form a fully enclosed shooting environment.

[0054] In other embodiments, the conveying unit can directly replace the support plate as part of the culture rack and be fixedly connected to the column. In this embodiment, the conveying unit serves both to support the culture tray and to convey the culture tray.

[0055] If the collection box has a covered structure, such as Figure 1 As shown, the acquisition box and the transmission unit together form a fully enclosed shooting environment.

[0056] Specifically, the culture rack can be a single-layer or multi-layer structure. If the culture rack is a multi-layer structure, the support plates or the conveying units that replace the support plates can be arranged vertically. The collection box is set on at least one layer of the culture rack.

[0057] like Figure 4 As shown, the four adjacent columns of the culture rack form the frame of the collection box, and the base plates between the columns enclose the collection box. If the culture rack has a multi-layer structure, the height of the collection box is less than or equal to the distance between two adjacent support plates in the vertical direction.

[0058] By setting up a collection box on the culture rack, the cost of the collection box is saved, and the culture rack provides support for the collection box, thus improving the stability of the collection box.

[0059] Furthermore, the image acquisition unit is a near-infrared camera, and at least some of the image acquisition units are configured with narrow-band filters whose center band is the preset band; preferably, the acquisition box also includes a light source assembly for providing a full-spectrum light source and an infrared light source; more preferably, the light source assembly is located on the upper part of the acquisition box and distributed around the near-infrared camera assembly.

[0060] In a preferred embodiment, the image acquisition unit is positioned at the center of the top wall of the acquisition box, so that the central axis of the camera lens is perpendicular to the shooting plane and passes through the geometric center of the shooting area. This ensures distortion-free viewing angle and full coverage of the shooting range, thereby guaranteeing image integrity.

[0061] Specifically, a camera mounting hole is provided in the center of the top wall of the data acquisition box, and the camera is fixed in the center of the top wall of the data acquisition box through the camera mounting hole.

[0062] Specifically, the image acquisition unit uses an 8MP resolution industrial near-infrared camera with a wavelength range of 200-1100nm to improve imaging quality.

[0063] In addition, the camera is equipped with a 10-50mm zoom lens and a high aperture to ensure that plant details are captured.

[0064] Specifically, if there is only one image acquisition unit used to acquire grayscale images in a preset band, then the image acquisition unit is equipped with multiple narrow-band filters with the center band as the preset band. By switching the narrow-band filters, multiple grayscale images corresponding to the preset bands can be captured.

[0065] If there are multiple image acquisition units used to acquire grayscale images in a preset band, each image acquisition unit is configured with a narrow-band filter with the preset band as the center band. In this way, each camera only needs to focus on acquiring grayscale images in one preset band.

[0066] In this embodiment, the plant image acquisition device also includes a driving component for driving multiple image acquisition units to take pictures simultaneously, avoiding time differences in the images taken by multiple cameras, which could lead to errors in the calculation of quality evaluation standards such as plant chlorophyll content.

[0067] Specifically, by configuring a narrow-band filter with the center band set to the preset band, only light of the preset band is allowed to pass through and reach the camera's image sensor, while light of other bands is significantly attenuated or blocked, thereby enabling the acquired grayscale image to accurately reflect the chlorophyll content-related signals.

[0068] The preset wavelengths include 550nm, 650nm, 710nm, 800nm, and 980nm. In addition, before acquiring images, parameters such as exposure time, white balance, and ISO are adjusted using professional software to improve the quality of images acquired by the camera in different environments.

[0069] like Figure 5 As shown, in this embodiment, there are two full-spectrum light sources, which are set opposite each other on the acquisition boxes on both sides of the camera to provide visible light close to natural light and avoid color shift; there are four infrared light sources, which are set at the four corners of the acquisition box to provide infrared light.

[0070] In a preferred embodiment, both the light source assembly and the image acquisition unit are disposed on the top wall of the acquisition line.

[0071] In one embodiment, the full-spectrum light source is a strip-shaped LED lamp, and the infrared light source is a tungsten filament lamp with a power of 120W.

[0072] Specifically, a diffuser is installed within the light source, meaning the diffuser is integrated into the light source assembly. This achieves pre-uniformization and orientation optimization of the incident light, eliminating glare, bright spots, and illumination gradients at the light emission source. This provides a highly uniform and stable illumination environment for image acquisition, plant phenotyping, and ultimately ensures the quantization reliability of grayscale images.

[0073] This invention also provides a plant image acquisition system, including the aforementioned plant image acquisition device, and further comprising: a control module for controlling the plant image acquisition device to simultaneously capture color images and grayscale images of plants in a target culture tray under a preset wavelength; and a processing module for determining, based on the images captured by the plant image acquisition device, the plant index score, plant weight, and plant phenotypic ratio of the plants in the target culture tray; and determining the quality grade of the plants in the target culture tray based on at least one of the plant index score, plant weight, and target plant phenotypic ratio.

[0074] Specifically, the plant index is a quantitative indicator based on image analysis, used to reflect the physiological state of plants, such as chlorophyll content and photosynthetic efficiency, thereby determining whether plants are healthy, have the potential for good growth, yield, and nutrient accumulation, and thus assessing plant quality.

[0075] For example, a high chlorophyll content in a plant usually indicates that the plant is fresher and more nutritious, thus assessing the plant as being of excellent quality.

[0076] Among them, NDVI is the Normalized Difference Vegetation Index, GNDVI is the Normalized Difference Vegetation Index for Green Light, NDGI is the Normalized Difference Vegetation Index for Greenness, CIG is the Chlorophyll Index for Green Light, GLI is the Green Leaf Index, MCARI is the Modified Chlorophyll Absorption and Reflectance Index, and MTVI2 is the Modified Triangular Vegetation Index 2.

[0077] The above seven plant indices reflect the physiological state of forage grass from dimensions such as vegetation cover, biomass, chlorophyll content, chlorophyll activity, water status, photosynthetic efficiency, leaf integrity, and nutrient level.

[0078] In this embodiment, the plant index score is the average of at least two plant indices.

[0079] Specifically, plant weight reflects the total amount of organic matter actually accumulated by the plant, and is a direct reflection of resource utilization efficiency and growth quality.

[0080] In this embodiment, the plant phenotype includes at least one of healthy, mild alopecia areata, severe alopecia areata, and fungal infection. The plant phenotype is determined by the observable characteristics exhibited by the plant and is a comprehensive external manifestation of plant quality. It includes growth potential, final outcome, and also covers dimensions that cannot be directly reflected by weight and indices, such as stress resistance and commercial quality.

[0081] Observable characteristics include plant height, leaf area, leaf color, fruit set rate, and fruit size.

[0082] The three types of quality assessment data mentioned above indirectly or partially reflect plant quality. However, relying solely on a single indicator can easily lead to misjudgments. By assessing the quality level of plants based on at least one of the following: plant index score, plant weight, and the proportion of the target plant phenotype, the coverage of assessment dimensions can be improved, a more comprehensive quality assessment standard can be constructed, and thus the accuracy of the results can be enhanced.

[0083] In one embodiment, the control module is further configured to send control commands to the plant image acquisition device, which in turn sends the control commands to the drive component, thereby enabling the drive component to control multiple image acquisition units to capture images at the same time.

[0084] The control commands are acquired or generated by the control module and are used to control the plant image acquisition device to capture color images and grayscale images of the plants in the target culture tray at the same time.

[0085] In one embodiment, after the image is captured, the plant image acquisition device is further used to process the image to obtain an image data stream, and transmit the image data stream to the processing module through a high-speed interface, so that the processing module receives and parses the data stream to obtain the captured image or the image data of the captured image, thus ensuring data integrity.

[0086] It should be noted that the target culture tray is the culture tray that enters the collection box. The target culture tray can be identified by laser detection, image recognition, user specification, or other methods.

[0087] Furthermore, the processing module is also used to extract image features from the image and calculate a plant index score and / or plant weight based on the extracted image features; preferably, the plant weight is calculated based on image features of a preset ratio and a specific preset band; preferably, the plant index score is calculated based on image features of a color image and / or a grayscale image under a preset band using a preset formula; more preferably, the plant index includes at least two of the following: NDVI, GNDVI, NDGI, CIG, GLI, MCARI, MTVI2.

[0088] In a preferred embodiment, the specific preset wavelength is 710nm, the image feature is the average pixel value, and the preset ratio is the ratio of the average pixel value of the 710nm wavelength image to the plant weight of 0.71. It should be noted that different plant species will have different specific preset wavelengths and preset ratios.

[0089] In this embodiment, before calculating the plant weight, a CNN regression model is trained with a grayscale image in the 710nm band as input and the plant weight as output, which is used to achieve non-contact plant weight prediction.

[0090] In another preferred embodiment, the preset formula includes the calculation formula for each plant index and the calculation formula for the average value of the plant indices.

[0091] Based on the above, the plant index score is calculated based on the average of at least two plant indices.

[0092] As an example, the calculation formula for NDVI includes the calculation formula for the NDVI index value corresponding to each pixel value, the calculation formula for at least one statistical analysis indicator, and the calculation formula for the normalization of the statistical analysis indicator.

[0093] Specifically, the NDVI index value is calculated for each pixel value in the image, and then a statistical analysis index is calculated based on the NDVI index value corresponding to each pixel value. Finally, the statistical analysis index is normalized to obtain the NDVI.

[0094] The statistical analysis indicators include at least one of the following: mean, standard deviation, coefficient of variation, 25th percentile, and 75th percentile. The normalization process includes obtaining a weighted average based on the above statistical analysis indicators and preset weights.

[0095] In one embodiment, the weight corresponding to the mean is 0.7, the weight corresponding to the standard deviation is 0.7, the weight corresponding to the coefficient of variation is 0.05, the weight corresponding to the 25th percentile is 0.1, and the weight corresponding to the 75th percentile is 0.1.

[0096] Other plant index calculation methods follow the same direction as NDVI calculation.

[0097] Furthermore, the processing module is also used to segment the image based on the trained image segmentation model, wherein the segmented image includes at least two phenotypic feature regions; and to determine the phenotypic proportion of the target plant based on the area of ​​the phenotypic feature regions; preferably, the phenotypic feature regions include healthy regions, mild alopecia areata regions, severe alopecia areata regions, and fungal infection regions.

[0098] Specifically, the training process of the image segmentation model includes: S1, creating the model; S2, acquiring data; S3, image processing; and S4, training the model.

[0099] Specifically, S2, data acquisition includes: identifying plant samples and taking images and color images of the samples in specific wavelength bands.

[0100] Specifically, S3 image processing includes: preprocessing, annotation, image enhancement, and dataset partitioning.

[0101] Preprocessing includes steps such as color temperature correction and cropping of the sample image; The annotation process includes labeling different plant phenotypic regions in the preprocessed image; the annotation process may include manual annotation and / or machine annotation, and manual annotation requires cross-validation to review and revise images with problems.

[0102] Image enhancement includes rotation, horizontal and vertical flipping, random scaling, random adjustment of brightness and contrast, addition of Gaussian noise and blur, adjustment of hue, saturation or brightness, and generation of multiple enhanced samples.

[0103] In addition, after the training process is completed, there is also a verification process for the image segmentation model, which is used to verify the model's functional stability and the accuracy of the results.

[0104] Furthermore, it also includes a display interaction module for receiving user input; the processing module is also used to determine the target culture tray based on the user input; preferably, the display interaction module is also used to display the quality grade of the plant in the target culture tray and / or the segmented image; more preferably, it also includes a storage module for storing the quality grade of the plant in the target culture tray and / or the image; the processing module is also used to determine the plant information to be photographed and the camera information for photographing the plant based on the user input; and to name the captured image based on the plant information and the camera information.

[0105] Specifically, user input includes camera information input, plant information input of the plant to be photographed, and control commands to start shooting.

[0106] Specifically, the images displayed in the interactive display module are either photographed plant images or processed plant images.

[0107] Specifically, plant information includes plant location information, seed type, seed source, germination date, shelf placement date, shipping date, serial number, weight, etc.; plant location information may include shelf number and culture tray number.

[0108] Specifically, the camera information used to photograph the plant is determined based on the selected camera and its parameter adjustments. This camera information includes the camera ID, exposure time, white balance, and ISO. Users can adjust these camera parameters by inputting this information to ensure optimal image quality under various environmental conditions.

[0109] During the camera parameter adjustment process, the display and interaction module is also used to display real-time preview images under different parameters.

[0110] Specifically, a preset image naming rule is established. After determining the target culture tray, the system automatically obtains the plant and camera information corresponding to the target culture tray and names the image based on the plant and camera information. As an example, the preset image naming rule is "[Photo Number][Shelf Number][Seed Source][Weight][Camera Number]", for example, "0_A-6-1-2-2_7.15_2.jpg".

[0111] The naming process described above can be performed by the processing module or the plant image acquisition device.

[0112] It should be noted that the images stored by the storage module are named images, and the storage module also stores the quality level of the plants in the target culture tray and the named images accordingly to form an instruction evaluation record.

[0113] like Figure 6 As shown, this embodiment of the invention also provides a plant quality monitoring method, comprising: simultaneously capturing a color image and a grayscale image of a plant in a target culture tray under a preset wavelength; determining a plant index, plant weight, and plant phenotypic percentage corresponding to the plant in the target culture tray based on the captured images; and determining the quality grade of the plant in the target culture tray based on at least one of the plant index, plant weight, and target plant phenotypic percentage.

[0114] In one embodiment, the method further includes: extracting image features from the image and calculating a plant index score and / or plant weight based on the extracted image features; segmenting the image based on a trained image segmentation model, wherein the segmented image includes at least two phenotypic feature regions; and determining the phenotypic proportion of the target plant based on the area of ​​the phenotypic feature regions.

[0115] In one embodiment, calculating a plant index score and / or plant weight based on extracted image features includes: calculating plant weight based on image features of a preset ratio and a specific preset band; and calculating a plant index score based on image features of a color image and / or a grayscale image under a preset band, using a preset formula.

[0116] In one embodiment, the method further includes: receiving user input; determining, based on the user input, the target culture tray, information about the plant to be photographed, and camera information for photographing the plant; photographing a color image and a grayscale image of the plant in the target culture tray under a preset wavelength; naming the photographed images based on the plant information and camera information; determining, based on the images photographed by the plant image acquisition device, the plant index score, plant weight, and plant phenotypic ratio corresponding to the plant in the target culture tray; and determining the quality grade of the plant in the target culture tray based on at least one of the plant index score, plant weight, and target plant phenotypic ratio; storing the quality grade of the plant in the target culture tray and / or the named image; and displaying the quality grade of the plant in the target culture tray and / or the segmented image.

[0117] In summary, the present invention provides a plant image acquisition device, a plant quality monitoring system, and a method, which have the following advantages compared with the prior art: The plant image acquisition device, composed of a collection box and a conveyor unit, eliminates the need for manual handling of culture trays from the conveyor belt to the collection box, removing intermediate transfer links and significantly improving processing efficiency. It upgrades the single acquisition function into a core component of industrial production, adapting to the needs of large-scale production and supporting production line linkage and intelligent decision-making. The use of the culture rack column support as the framework of the collection box also saves costs. The closable opening on the collection box reduces external light interference, providing a high-quality data source for spectral analysis and phenotypic segmentation. Phenotypic segmentation, weight prediction, and plant indices provide data support for a comprehensive quantitative assessment of plant quality. Based on at least one of the following: plant index score, plant weight, and the proportion of the target plant phenotype, the quality level of the plant in the target culture tray is determined, constructing a more comprehensive quality assessment standard, improving the coverage of assessment dimensions, thereby increasing the accuracy of results and overcoming the problem of misjudgment that can easily occur when relying on a single indicator.

[0118] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-described technical content to create equivalent embodiments without departing from the scope of the present invention. The implementation schemes in the above embodiments can also be further combined or replaced. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A plant image acquisition device, characterized by, include: An image acquisition box, wherein an image acquisition unit is installed inside the image acquisition box; A conveying unit used to transport plant culture trays; The acquisition box is provided with a closable opening, which is used to dock with the transmission unit so that the image acquisition unit can acquire images of the culture trays transported into the acquisition box.

2. The plant image acquisition device according to claim 1, characterized in that, The opening includes an inlet and an outlet; The conveying unit delivers the culture tray into the collection box through the inlet and sends the culture tray out of the collection box through the outlet. The image acquisition unit is located on the upper part of the acquisition box; Preferably, the number of image acquisition units is at least two, which are used to acquire color images of plants and grayscale images under preset bands, respectively.

3. The plant image acquisition device according to claim 1 or 2, characterized in that, The acquisition box includes a box frame and a base plate; The box frame is located above the conveying unit; The substrate surrounds at least part of the housing frame and forms a collection box above the transmission unit.

4. The plant image acquisition apparatus according to claim 3, characterized by It also includes, A culture rack, wherein at least a portion of the culture rack is provided with a conveying unit; The culture rack includes a column, which intersects the conveying unit in the vertical direction; At least some of the columns are used to form the frame of the collection box.

5. The plant image acquisition device according to any one of claims 1-4, characterized in that, The image acquisition unit is a near-infrared camera, and at least some of the image acquisition units are equipped with narrow-band filters whose center band is the preset band; Preferably, the acquisition box also includes a light source assembly for providing a full-spectrum light source and an infrared light source; More preferably, the light source assembly is disposed on the upper part of the acquisition box and distributed around the near-infrared camera assembly.

6. A plant quality monitoring system comprising the plant image capturing device according to any one of claims 1 to 5, characterized in that, Also includes: The control module is used to control the plant image acquisition device to capture color images and grayscale images of the plants in the target culture tray at the same time. The processing module is used to determine the plant index score, plant weight, and plant phenotypic ratio of the plant in the target culture tray based on the image captured by the plant image acquisition device; and to determine the quality grade of the plant in the target culture tray based on at least one of the plant index score, plant weight, and target plant phenotypic ratio.

7. The plant quality monitoring system of claim 6, wherein, The processing module is also used for, Extract image features from the image, and calculate the plant index score and / or plant weight based on the extracted image features; Preferably, the plant weight is calculated based on a preset ratio and image features of a specific preset band; Preferably, the plant index score is calculated based on a preset formula, according to the image features of a color image and / or a grayscale image under a preset band; More preferably, the plant index includes at least two of the following: NDVI, GNDVI, NDGI, CIG, GLI, MCARI, and MTVI2.

8. The plant quality monitoring system according to claim 6 or 7, characterized in that The processing module is also used for, The image is segmented based on a trained image segmentation model, and the segmented image includes at least two phenotypic feature regions. The proportion of the target plant phenotype is determined based on the area of ​​the phenotypic feature region. Preferably, the phenotypic feature regions include healthy regions, mild alopecia areata regions, severe alopecia areata regions, and fungal infection regions.

9. The plant quality monitoring system of claim 8, wherein, It also includes, The display interaction module is used to receive user input; The processing module is also used to determine the target culture tray based on the user input; Preferably, the display interaction module is further configured to display the quality grade of the plants in the target culture tray and / or the segmented image; More preferably, it also includes a storage module for storing the quality level and / or image of the plants in the target culture tray; The processing module is further configured to determine the plant information to be photographed and the camera information for photographing the plant based on the user input; and to name the captured image based on the plant information and the camera information.

10. A plant quality monitoring method applied to the plant quality monitoring system according to any one of claims 6 to 9, characterized by, include: Capture color images and grayscale images of the plants in the target culture tray at the same time; Based on the captured images, determine the plant index, plant weight, and plant phenotypic percentage corresponding to the plants in the target culture tray; The quality grade of the plants in the target culture tray is determined based on at least one of the plant index, plant weight, and target plant phenotypic percentage.

Images