Soilless culture crop assembly line growth monitoring system and method

Abstract

The invention discloses a soilless culture crop assembly line growth monitoring system and method, and the system comprises an image collection module, a control module and a grass disc conveying belt, and the image collection module is in communication connection with the control module. The image acquisition module is arranged above the grass disc, when the sensor detects the grass disc, a trigger signal is sent out, the number input device reads the number of the grass disc, the image acquisition module acquires an image of the grass disc, the processor performs spectral feature extraction and registration on the acquired image, and a crop growth state index is calculated based on a vegetation index model. Spectral information of a plurality of wave bands can be simultaneously acquired in one-time image acquisition, the information acquisition efficiency is high, the growth state of crops in a grass tray can be quickly identified, the response speed is high, and the method is suitable for a grass tower dense space and a scene of quickly detecting the growth state of the crops in an assembly line.

Description

Technical Field

[0001] This invention belongs to the field of intelligent agriculture technology, specifically, it relates to a hydroponic growth monitoring system and method for water lines in soilless cultivation. Background Technology

[0002] Soilless cultivation offers advantages such as space saving, greenhouse cultivation, and high resource utilization, making it widely used in agricultural production. To ensure the growth of soilless crops, it is necessary to regularly monitor indicators such as leaf morphology and chlorophyll content.

[0003] Previously, hyperspectral cameras were used to monitor crop growth. However, the high planting density of grass towers, with interlayer heights of approximately 300mm, makes hyperspectral cameras bulky due to their complex spectroscopic components. Furthermore, they require a high working distance (at least 500mm), making them unsuitable for use between grass tower layers. Additionally, hyperspectral images generate large amounts of data, and data transmission and processing can take several hours, which is incompatible with the millisecond-level response time of automated grass tower production lines. Moreover, the hydroponic environment inside the grass towers causes significant temperature and humidity fluctuations. Hyperspectral spectroscopic components are highly precise, and large temperature and humidity changes can affect their accuracy, leading to inaccurate readings. Frequent calibration also disrupts the production process.

[0004] To address these issues, existing technologies employ color RGB cameras for imaging, which are low-cost, compact, and easy to deploy. However, the acquired spectrum is singular, only capable of detecting morphological characteristics of crops and unable to quantify core indicators such as chlorophyll and water stress, making comprehensive crop growth assessment difficult. Furthermore, there is no optimization for the interlayer height of the grass towers, resulting in insufficient field-of-view coverage and missed detections at edges.

[0005] In view of this, the present invention is proposed. Summary of the Invention

[0006] The technical problem this invention aims to solve is to overcome the shortcomings of existing technologies. The first aspect of this invention provides a hydroponic crop growth monitoring system that can achieve ultra-close-range wide-angle coverage, comprehensively assess crop growth status, and has a fast response speed. This solves the problems of missed detections, insufficient imaging band coverage, and slow response speed in existing technologies when used between straw tower layers.

[0007] The second aspect provides a method for monitoring the growth of hydroponically grown crops in water lines.

[0008] To solve the above-mentioned technical problems, the basic concept of the technical solution adopted by the present invention is as follows: A hydroponic growth monitoring system for soilless cultivation includes: Image acquisition module: used to acquire images of soilless cultivation crop trays; Control module: Controls the image acquisition unit to acquire images of the grass tray, and processes and analyzes the images; The image acquisition module includes color and near-infrared binocular cameras with their optical axes set parallel to each other.

[0009] Furthermore, it includes a grass tray conveyor belt, on which the grass tray is disposed and transported along a set direction, and the binocular camera is fixed above the grass tray.

[0010] Furthermore, the grass tray conveyor belt is set on several layers of three-dimensional circulating grass towers, and the binocular camera is set between the layers of the grass towers. The color camera of the binocular camera is equipped with a color wide-angle lens, and the near-infrared camera is equipped with a near-infrared narrow-band wide-angle lens.

[0011] Furthermore, the image acquisition module includes a supplementary light, which is positioned above the grass tray conveyor belt for supplementary lighting; Preferably, a supplementary lighting bracket is connected to the grass tower, and the supplementary lighting is installed on the extension of the supplementary lighting bracket toward the conveyor belt; More preferably, the binocular camera is mounted on the extension of the fill light bracket to the conveyor belt; Preferably, the supplementary light is integrally formed with the bracket extending into the conveyor belt; Preferably, the supplementary light comprises alternating 450nm blue LEDs, 540nm green LEDs, 660nm red LEDs, and 850nm near-infrared LEDs, with the same number of each of the four types of LEDs.

[0012] Furthermore, the control module includes a sensor, which is set at a first preset position outside the grass tray conveyor belt to detect the position of the grass tray.

[0013] Furthermore, the control module includes a number input device, which is set at a second preset position outside the grass tray conveyor belt and is used to input the grass tray number; The direction from the second preset position to the first preset position is consistent with the direction of grass tray transmission.

[0014] Furthermore, the control module includes a processor. The processor extracts feature points from the image and aligns the color image with the near-infrared image pixels based on these feature points. Plant regions in images are extracted using an automatic thresholding segmentation algorithm; Extract the spectral feature vectors of each pixel in the plant region in the four bands of red, green, blue and near-infrared. By inputting the spectral feature vector of a pixel into at least one of the NDVI, GNDVI, CIG, EVI, GRVI, NDWI and PSRI vegetation index models, a plant index is obtained. The mean and variance of the plant index are calculated to obtain an index reflecting the growth status of soilless cultivated crops.

[0015] Furthermore, the process of aligning the color image and the near-infrared image pixels based on feature points is as follows: The LOFTR algorithm based on the Transformer architecture is used to match feature points between the color image and the near-infrared image to obtain matching point pairs; Based on the matching point pairs, the homography matrix from the near-infrared image to the color image is estimated using the USAC_ACCURATE strategy, and the reprojection error of the homography matrix is ​​no greater than 1 pixel. Using the color image as a reference, the near-infrared image is transformed using the homography matrix to achieve pixel alignment between the near-infrared image and the color image.

[0016] Furthermore, the control module includes a memory for storing images acquired by the image acquisition unit and the growth status index of hydroponically grown crops obtained by the processor.

[0017] A method for monitoring the growth of hydroponically grown crops in a hydroponic system includes the following steps: S1. Perform system initialization and calibration; S2. When soilless cultivation crop trays are detected, collect the tray number, color image, and near-infrared image. S3. Align the pixels of the collected color and near-infrared images of the grass tray to delineate the plant areas in the images; S4. Extract the spectral feature vector of each pixel in the plant region; S5. Substitute the extracted spectral feature vector into at least one of the NDVI, GNDVI, CIG, EVI, GRVI, NDWI and PSRI vegetation index models to obtain the plant index. S6. Calculate the crop growth status index based on the plant index; S7. The grass tray number, image, and crop growth status index are associated and stored.

[0018] By adopting the above technical solution, the present invention has the following beneficial effects compared with the prior art: This invention achieves a total latency of less than 1 second from image acquisition to exponential output, meeting the second-level response production rhythm of the hydroponic system. The binocular camera can simultaneously acquire RGB and near-infrared information in a single acquisition, significantly reducing costs compared to hyperspectral cameras. It can also adapt to the close interlayer spacing of the hydroponic system, achieving ultra-close-range wide-angle coverage without missing any detections. This provides an efficient and reliable automated monitoring method for precision agriculture in hydroponic environments.

[0019] The monitoring system works in conjunction with the conveyor belt, enabling the binocular camera to capture clear images as the conveyor belt moves, thus improving detection efficiency.

[0020] Setting a wide-angle lens enables close-up imaging, avoids insufficient field of view coverage, and is suitable for capturing images between layers of grass towers.

[0021] The supplementary light provides four bands of illumination for the binocular camera, ensuring the quality of the acquired images and laying the foundation for analysis.

[0022] By detecting the position of the grass tray using sensors, the system can determine the timing of image acquisition based on the sensor's trigger signal, thus achieving automated acquisition.

[0023] By introducing a numbered data entry unit and linking the grass tray number with image data and crop growth status, precise traceability and management of production data are achieved. Operators can quickly retrieve all historical data for any grass tray using its number, improving production management and troubleshooting efficiency.

[0024] By employing pixel-level image registration, multi-band spectral feature vector extraction, and calculation of multiple vegetation indices, a comprehensive assessment of the growth status of soilless cultivated crops is achieved with high accuracy.

[0025] By using the LOFTR algorithm based on the Transformer architecture for feature matching and the USAC_ACCURATE strategy to calculate the homography matrix, the registration problem of plant images is effectively solved, providing a foundation for subsequent multispectral information fusion and plant index calculation.

[0026] The specific embodiments of the present invention will now be described in further detail with reference to the accompanying drawings. Attached Figure Description

[0027] The accompanying drawings, as part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments and descriptions of the invention are used to explain the invention, but do not constitute an undue limitation of the invention. Obviously, the drawings described below are merely some embodiments, and those skilled in the art can obtain other drawings based on these drawings without creative effort. In the drawings: Figure 1 This is a diagram showing the layout of the photographing location equipment for the grass tower transmission line of the present invention; Figure 2 This is a side view of the layout diagram of the photographing location device for the grass tower transmission line of the present invention; Figure 3 This is a schematic diagram of the LED arrangement of the supplementary light of the present invention; Figure 4 This is a flowchart of the present invention.

[0028] 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

[0029] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments will be clearly and completely described below with reference to the accompanying drawings. The following embodiments are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0030] In the description of this invention, it should be noted that the terms "upper", "lower", "front", "rear", "left", "right", "vertical", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting this invention.

[0031] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0032] A hydroponic growth monitoring system for water lines in soilless cultivation includes an image acquisition module and a control module with communication connections, as well as a grass tray conveyor belt.

[0033] The soilless cultivation crops include leafy plants suitable for the grass tower automated hydroponics technology, such as leafy vegetables and sprouts.

[0034] The grass trays are placed on a conveyor belt and transported along a set direction. The grass tray conveyor belt is set on several layers of three-dimensional circulating grass towers. The grass tower track is a multi-layered closed loop. The slope of each layer of the grass tower is small and the space between layers is compact.

[0035] The image acquisition module consists of a binocular camera and a supplementary light. The binocular camera includes a color camera and a monochrome near-infrared camera. The color camera is equipped with a color wide-angle lens, and the monochrome near-infrared camera is equipped with an 850nm narrowband wide-angle lens. The binocular camera is mounted on the extension of the supplementary light bracket onto the conveyor belt, with its optical axis parallel to the light. The camera resolution is set to 1280×720 pixels, the frame rate is set to 30fps, and the average exposure time is 30ms. The binocular camera assembly is installed 320±10mm above the grass tray conveyor line, equidistant from both sides of the conveyor belt, with the two cameras arranged along the long side of the grass tray. Figure 2 and Figure 3 As shown, the binocular camera has a field of view of 800mm × 500mm. The camera supports the UVC protocol and connects to the control module via USB 3.0. One image acquisition module is installed on each floor of the grass tower, or one image acquisition module is installed every other floor.

[0036] The supplementary light is configured as an 800mm × 60mm strip, each light comprising alternating 450nm blue LEDs, 540nm green LEDs, 660nm red LEDs, and 850nm near-infrared LEDs, with the same number of each type. The LED arrangement is as follows: Figure 3 As shown, the supplementary light bracket is connected to the grass tower, and the supplementary light is set on the extension of the supplementary light bracket to the conveyor belt, at a height of 320±10mm above the grass tray transmission line, to provide sufficient multi-band supplementary light for the binocular camera assembly, thereby eliminating ambient light interference and ensuring that the image reflectivity fluctuation is ≤5%.

[0037] The control module controls the image acquisition unit to acquire images of the grass tray, and processes and analyzes the images. The control module includes a sensor, a numbering input device, a processor, and a memory.

[0038] The sensor has a triggering accuracy of ≤1ms and is positioned at a first preset position outside the grass tray conveyor belt, such as... Figure 1 and Figure 2 As shown. The orthographic projection point of the binocular camera in the direction of the conveyor belt is taken as the reference point, and the first preset position is 250mm downstream of the reference point.

[0039] The number input device is positioned at a second preset position outside the grass tray transmission line. The direction of the second preset position toward the first preset position is consistent with the grass tray transmission direction. The position is as follows: Figure 1 and Figure 2 As shown. The second preset position is the orthographic projection point of the binocular camera in the direction of the conveyor belt's movement. It is used to input the number of the grass trays. The number input device connects to the control module via USB 2.0 and has IP65 industrial-grade dust and water resistance.

[0040] The processor performs reflectance correction, spectral feature extraction and registration, plant region segmentation, and plant index analysis on the acquired images based on reflectance correction parameters.

[0041] The memory is used to store images acquired by the image acquisition unit and growth status indices of hydroponically grown crops obtained by the processor. It can be stored locally or in the cloud.

[0042] The system's working process is as follows: Figure 4 As shown, the details are as follows: S101. Perform system initialization and calibration.

[0043] After the system is powered on, the control module loads the parameter configuration file of the binocular camera and initializes the camera's exposure time, resolution, and frame rate.

[0044] The control module acquires reflectivity correction parameters. With the binocular camera and supplementary light on, it captures a five-level grayscale image. The control module extracts the brightness values ​​of each grayscale region in the image, compares them with standard brightness values, and uses a quadratic polynomial algorithm for fitting, for example, y = 0.001x² + 0.92x + 5. The reflectivity correction parameters (a=0.001, b=0.92, c=5) are obtained and saved locally for automatic correction of all subsequent acquired images.

[0045] S102, Data Acquisition.

[0046] The grass tower conveyor belt starts operating. Since the binocular camera's field of view is 800mm × 500mm and its resolution is 1280 × 720 pixels, 500mm ÷ 720 pixels ≈ 0.694mm / pixel. The camera's exposure time is 30ms, so 0.694mm ÷ 0.03s ≈ 23mm / s. Therefore, to ensure imaging quality, the conveyor belt's speed must be ≤23mm / s.

[0047] When the sensor detects a grass tray, it sends a trigger signal. The numbering device scans the QR code on the grass tray to obtain its number, and the processor verifies the number format. The control module controls the binocular camera to acquire images. The binocular camera simultaneously acquires color and near-infrared images, which are then transmitted to the control module via a USB 3.0 interface.

[0048] S103, Data Preprocessing.

[0049] The processor performs reflectivity correction on the acquired image using the reflectivity correction coefficients obtained by S101 to eliminate uneven lighting and lens effects.

[0050] S104, Data Analysis.

[0051] The processor registers the preprocessed image and uses the LOFTR algorithm based on the Transformer architecture to match feature points between the color image and the near-infrared image to obtain matching point pairs. Based on the matching point pairs, the homography matrix from the near-infrared image to the color image is estimated using the USAC_ACCURATE strategy, and the reprojection error of the homography matrix is ​​no greater than 1 pixel. Using the color image as a reference, the near-infrared image is transformed using the homography matrix to achieve pixel alignment between the near-infrared image and the color image; The plant region in the aligned image was extracted using an automatic thresholding segmentation algorithm. Extract the spectral feature vectors of each pixel in the plant region in the four bands of red, green, blue and near-infrared. By inputting the spectral feature vector of a pixel into at least one of the NDVI, GNDVI, CIG, EVI, GRVI, NDWI and PSRI vegetation index models, a plant index is obtained. Calculate the mean and variance of the plant index to obtain an index that reflects the crop growth status; The calculated index results are summarized to generate a crop growth status report.

[0052] Furthermore, the process of aligning the color image and the near-infrared image pixels based on feature points is as follows: The LOFTR algorithm based on the Transformer architecture is used to match feature points between the color image and the near-infrared image to obtain matching point pairs; Based on the matching point pairs, the homography matrix from the near-infrared image to the color image is estimated using the USAC_ACCURATE strategy, and the reprojection error of the homography matrix is ​​no greater than 1 pixel. Using the color image as a reference, the near-infrared image is transformed using the homography matrix to achieve pixel alignment between the near-infrared image and the color image.

[0053] Furthermore, the calculation methods for the NDVI, GNDVI, CIG, EVI, GRVI, NDWI, and PSRI vegetation index models are as follows: NDVI (Normalized Difference Vegetation Index): (850nm-660nm) / (850nm+660nm); GNDVI (Green Normalized Difference Vegetation Index): (850nm-540nm) / (850nm+540nm); CIG (Green Chlorophyll Index): 850nm / 540nm-1; EVI (Enhanced Vegetation Index): ((850nm-660nm) / (850nm+6×660nm-7.5×450nm+1))×2.5; GRVI (Green-Red Vegetation Index): (540nm-660nm) / (540nm+660nm); NDWI (Normalized Difference Water Index): (540nm-850nm) / (540nm+850nm); PSRI (Plant Senescence Reflectance Index): (660nm-450nm) / 850nm.

[0054] Here, 850nm represents the reflectance value of the near-infrared band, 660nm represents the reflectance value of red light, 540nm represents the reflectance value of green light, and 450nm represents the reflectance value of blue light.

[0055] S105, Data storage.

[0056] The processor associates the generated crop growth status report with the grass tray number and stores it in memory.

[0057] In another implementation, a historical data comparison function is added. A time-series interface is added to the cloud storage module, and software is added to implement historical data comparison. Through time-series analysis, crop growth trends are predicted, and an early warning function is provided to users.

[0058] In another implementation, a user interface is added, allowing users to view crop growth indices by scanning the QR code on the grass tray.

[0059] In another embodiment, this scheme is applied to other hydroponic crops, and the wavelength of the strip supplemental light is adjusted according to the wavelength of the detected plant index.

[0060] 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 hydroponic growth monitoring system for soilless cultivation, characterized in that, include: Image acquisition module: used to acquire images of soilless cultivation crop trays; Control module: Controls the image acquisition unit to acquire images of the grass tray, and processes and analyzes the images; The image acquisition module includes color and near-infrared binocular cameras with their optical axes set parallel to each other.

2. The hydroponic growth monitoring system for soilless cultivation according to claim 1, characterized in that, It includes a grass tray conveyor belt, on which grass trays are placed and transported in a set direction, and the binocular camera is fixed above the grass trays.

3. The hydroponic growth monitoring system for soilless cultivation according to claim 2, characterized in that, The grass tray conveyor belt is set on several layers of three-dimensional circulating grass towers. The binocular camera is set between the layers of the grass towers. The color camera of the binocular camera is equipped with a color wide-angle lens, and the near-infrared camera is equipped with a near-infrared narrow-band wide-angle lens.

4. The hydroponic growth monitoring system for soilless cultivation according to claim 1, characterized in that, The image acquisition module includes a supplementary light, which is positioned above the grass tray conveyor belt for supplementary lighting; Preferably, a supplementary lighting bracket is connected to the grass tower, and the supplementary lighting is installed on the extension of the supplementary lighting bracket toward the conveyor belt; More preferably, the binocular camera is mounted on the extension of the fill light bracket to the conveyor belt; Preferably, the supplementary light is integrally formed with the bracket extending into the conveyor belt; Preferably, the supplementary light comprises alternating 450nm blue LEDs, 540nm green LEDs, 660nm red LEDs, and 850nm near-infrared LEDs, with the same number of each of the four types of LEDs.

5. A hydroponic growth monitoring system for soilless cultivation according to claim 2, characterized in that, The control module includes a sensor, which is set at a first preset position outside the grass tray conveyor belt to detect the position of the grass tray.

6. The hydroponic growth monitoring system for soilless cultivation according to claim 5, characterized in that, The control module includes a number input device, which is set at a second preset position outside the grass tray conveyor belt and is used to input the grass tray number. The direction from the second preset position to the first preset position is consistent with the direction of grass tray transmission.

7. The hydroponic growth monitoring system for soilless cultivation according to claim 1, characterized in that, The control module includes a processor. The processor extracts feature points from the image and aligns the color image with the near-infrared image pixels based on these feature points. Plant regions in images are extracted using an automatic thresholding segmentation algorithm; Extract the spectral feature vectors of each pixel in the plant region in the four bands of red, green, blue and near-infrared. By inputting the spectral feature vector of a pixel into at least one of the NDVI, GNDVI, CIG, EVI, GRVI, NDWI and PSRI vegetation index models, a plant index is obtained. The mean and variance of the plant index are calculated to obtain an index reflecting the growth status of soilless cultivated crops.

8. The hydroponic growth monitoring system for soilless cultivation according to claim 7, characterized in that, The process of aligning the color image and the near-infrared image pixels based on feature points is as follows: The LOFTR algorithm based on the Transformer architecture is used to match feature points between the color image and the near-infrared image to obtain matching point pairs; Based on the matching point pairs, the homography matrix from the near-infrared image to the color image is estimated using the USAC_ACCURATE strategy, and the reprojection error of the homography matrix is ​​no greater than 1 pixel. Using the color image as a reference, the near-infrared image is transformed using the homography matrix to achieve pixel alignment between the near-infrared image and the color image.

9. A hydroponic growth monitoring system for soilless cultivation according to claim 1, characterized in that, The control module includes a memory for storing images acquired by the image acquisition unit and growth status indices of hydroponically grown crops obtained by the processor.

10. A method for monitoring the growth of hydroponically grown crops, characterized in that, Includes the following steps: S1. Perform system initialization and calibration; S2. When soilless cultivation crop trays are detected, collect the tray number, color image, and near-infrared image. S3. Align the pixels of the collected color and near-infrared images of the grass tray to delineate the plant areas in the images; S4. Extract the spectral feature vector of each pixel in the plant region; S5. Substitute the extracted spectral feature vector into at least one of the NDVI, GNDVI, CIG, EVI, GRVI, NDWI and PSRI vegetation index models to obtain the plant index. S6. Calculate the crop growth status index based on the plant index; S7. The grass tray number, image, and crop growth status index are associated and stored.

Images