Plant status and space detection method and system based on image recognition

Abstract

This invention relates to a plant status and vacancy detection method and system based on image recognition, belonging to the field of planting monitoring technology. It includes acquiring image information of the planting area; obtaining current vegetation characteristics; determining the current vegetation category based on the current vegetation characteristics; determining the current vegetation growth status and recording it in a vegetation information table; acquiring and outputting the coordinates of abnormal vegetation growth areas; identifying abnormal vegetation areas; acquiring image information of abnormal vegetation areas to determine vegetation branch characteristics; identifying branches touching the soil based on the vegetation branch characteristics; determining abnormal growth and rooting characteristics when branches touching the soil exist; outputting vacancy area signals when abnormal growth and rooting characteristics exist; and outputting vacancy area signals when no current vegetation characteristics exist. This invention avoids misjudgment of vacancy caused by overlapping branches and leaves, improving the accuracy of plant status monitoring and vacancy determination.

Description

Technical Field

[0001] This invention relates to the field of planting monitoring technology, and in particular to a method and system for detecting plant status and vacancy based on image recognition. Background Technology

[0002] In modern planting scenarios such as greenhouses and large-scale farmland, monitoring vegetation growth is a core element in ensuring crop yields and optimizing field management. With the development of smart agriculture technologies, image recognition technology has been widely applied to monitoring planting areas.

[0003] Monitoring vacant planting areas is a key prerequisite for achieving rational utilization of planting resources and precise replanting. Its core requirement is to accurately identify areas in the planting troughs where there is no effective living vegetation, so as to provide data support for subsequent replanting operations and optimization of planting layout.

[0004] Regarding the aforementioned technologies, in the process of identifying vegetation status and detecting vacancy in planting areas, they cannot effectively distinguish invalid plants such as branches or leaves scattered in the soil, making it difficult to determine real vacancy and resulting in inaccurate vacancy detection results in planting areas. Summary of the Invention

[0005] To address the impact of invalid plants such as branches and fallen leaves on the detection of vacancy areas in planting areas, this invention provides a method and system for plant status and vacancy detection based on image recognition.

[0006] In a first aspect, the present invention provides a method for detecting plant status and vacancy based on image recognition, employing the following technical solution: A method for plant status and vacancy detection based on image recognition, comprising: Step S1: Obtain image information of the planting area; Step S2: Analyze the image information of the planting area to obtain the current vegetation features; Step S20: If current vegetation features exist, determine the current vegetation category based on the current vegetation features; Step S21: Determine the current vegetation growth status based on the current vegetation category and current vegetation characteristics, and record it in the preset vegetation information table; Step S22: If the current vegetation growth state is abnormal, obtain the coordinates of the abnormal vegetation growth area and output them; Step S23: When the current vegetation category does not exist, identify the abnormal vegetation area; Step S24: Obtain image information of abnormal vegetation areas based on abnormal vegetation areas, and analyze the image information of abnormal vegetation areas to determine the characteristics of vegetation branches and trunks; Step S25: Determine the soil adhesion to the branches based on the characteristics of vegetation branches; Step S26: When there is soil adhering to the branches, determine the abnormal growth and rooting characteristics; Step S27: When abnormal growth and rooting characteristics are present, output the vacancy region signal; Step S3: When there are no current vegetation features, output the vacancy area signal.

[0007] By employing the above technical solution, the current vegetation characteristics are obtained by acquiring image information of the planting area. When vegetation characteristics exist, the vegetation category and growth status are determined and recorded; if an anomaly is detected, the coordinates of the abnormal area are output. If no vegetation category exists, the soil adhesion to the branches is determined; if soil adhesion to the branches is present, it is judged as abnormal rooting and a void signal is output. If no vegetation characteristics exist, a void signal is directly output. This improves the accuracy of identifying normal vegetation, abnormally growing vegetation, and void areas.

[0008] Optional, also includes: Step S40: When there are no vegetation branch features or no abnormal growth and rooting features, obtain bud and leaf features based on the image information of the abnormal vegetation area. Step S41: When there are no buds or branches, output the empty area signal; Step S42: When there are bud and leaf characteristics and vegetation branch and trunk characteristics, determine the growth transition connection area based on the bud and leaf characteristics and vegetation branch and trunk characteristics; Step S43: If there is no growth transition connection region, output the vacancy region signal; Step S44: If a growth transition connection region exists, output an abnormal vegetation region signal.

[0009] By adopting the above technical solution, when there are no vegetation branch features or abnormal growth and rooting features, the bud and leaf features are obtained. If the bud and leaf features are also absent, the empty area signal is directly output. When both bud and leaf features and vegetation branch features are present, the growth transition connection area between the two is determined. If the connection area is absent, it indicates that the bud and leaf and the branch may not be organic components of the same plant, such as scattered branches and leaves. In this case, the empty area signal is output. If the growth transition connection area exists, it indicates that the bud and leaf and the branch belong to the same plant, but there may be other abnormal growth conditions. Therefore, the abnormal vegetation area signal is output. This more precisely eliminates the interference of invalid plants such as isolated buds, leaves, or branches scattered in the planting area on the empty area detection, thus improving the accuracy of empty area judgment.

[0010] Optionally, a method for outputting a reasonable empty space area signal is also included, the method comprising: Step S45: When there is a vacancy signal, search for vegetation characteristics in adjacent areas; Step S46: Determine the vegetation growth trend of adjacent areas based on the vegetation characteristics of adjacent areas; Step S47: If the vegetation growth trend in the adjacent area is an expansion trend, output a reasonable vacancy area signal; Step S48: If the vegetation growth trend in the adjacent area is not the same as the surrounding expansion growth trend, output the empty area signal; Step S49: When there is no vegetation growth trend in adjacent areas, determine the adjacent vegetation category based on the vegetation characteristics of adjacent areas; Step S50: Search the preset vegetation feature database based on the adjacent vegetation categories to determine the growth trend of adjacent vegetation, and then proceed to steps S47 to S48.

[0011] By employing the above technical solution, when a vacancy signal exists, the vegetation characteristics of adjacent areas are searched to determine the growth trend of vegetation in those areas. When the vegetation in adjacent areas shows a trend of peripheral expansion, it means that the vacancy may be a reasonable space reserved during the natural growth of the vegetation, or a normal gap formed by the expansion of surrounding vegetation. In this case, a reasonable vacancy signal is output. If the growth trend of the vegetation in adjacent areas is not peripheral expansion, such as showing a trend of concentrated growth in a specific direction or growth stagnation, then the vacancy may be a real gap that needs attention, and a vacancy signal is output. When the growth trend cannot be directly determined from the vegetation characteristics of adjacent areas, the category of adjacent vegetation is first identified based on the vegetation characteristics of adjacent areas. Then, the typical growth trend of this type of vegetation is obtained by querying the vegetation characteristic database. Subsequently, the reasonable vacancy judgment steps are performed based on this trend. This approach combines the inherent growth characteristics of vegetation to more accurately identify reasonable vacancy, avoiding misjudging normal growth gaps as vacancy that needs to be filled, thus improving the practicality and reliability of the vacancy detection results.

[0012] Optionally, it also includes a method for outputting a reasonable vacancy area signal when there are no adjacent vegetation types, the method comprising: Step S450: Obtain the current example region; Step S451: Determine the empty space area and the number of empty space areas based on the current example area; Step S452: Determine the length of the empty space area based on the empty space area; Step S453: Determine the total length of vacant spaces based on the length of vacant areas, and determine the average length of vacant spaces based on the number of vacant areas. Step S454: Determine the deviation value of the occupied length based on the occupied length of the vacant area and the average occupied length of the vacant area; Step S455: If the deviation value of the occupied length falls within the preset reasonable deviation value range, output a reasonable empty space area signal.

[0013] By adopting the above technical solution, when there are no adjacent vegetation categories, it is impossible to determine whether the vacancy is reasonable by conventionally associating growth trends with vegetation categories. In this case, by obtaining the current example area and using it as a reference sample, the number of vacancy areas is first determined. Then, for each determined vacancy area, its occupied length is measured and determined. The occupied lengths of all vacancy areas are then summed to obtain the total occupied length of the vacancy. The average occupied length of the vacancy is calculated in combination with the number of vacancy areas. The deviation value of the occupied length of each vacancy area from the average length is also calculated. When the occupied length deviation value falls within the reasonable deviation range, it indicates that the occupied length of the vacancy area is within the normal fluctuation range and has good consistency with the occupied situation of other vacancy areas in the example area. It can be judged as a reasonable vacancy area, thus outputting a reasonable vacancy area signal. This achieves effective identification of reasonable vacancy areas based on the statistical characteristics of the vacancy itself in the region, even when there is a lack of reference to adjacent vegetation categories.

[0014] Optionally, it also includes a method for outputting a reasonable empty space area signal if the deviation value of the occupied length does not fall within a reasonable deviation value range, the method including: Step S4550: Determine the current empty space region based on the current example region; Step S4551: Obtain the left and right adjacent vegetation regions corresponding to the current empty space region; Step S4552: Obtain the left adjacent region corresponding to the left adjacent vegetation region; Step S4553: Obtain the right adjacent region corresponding to the right adjacent vegetation region; Step S4554: If both the left and right adjacent areas are empty areas, output a reasonable empty area signal.

[0015] By employing the aforementioned technical solution, the current vacant area within the current example area is clearly identified. Then, the left and right adjacent vegetation areas corresponding to this current vacant area are obtained. Next, the left adjacent area outside the left adjacent vegetation area and the right adjacent area outside the right adjacent vegetation area are further obtained. If, after detection, both the left and right adjacent areas are found to be vacant areas—meaning that the left and right sides of the current vacant area are also vacant, forming a continuous vacant area—in this case, even if the deviation in the occupied length of the current vacant area exceeds the preset range, it may be due to concentrated vacant areas existing in the planting plan. Therefore, it is determined to be a reasonable vacant area and a corresponding signal is output. This provides a more comprehensive consideration of the actual distribution of vacant areas in the planting area, avoiding misjudgments due to abnormal lengths of a single vacant area.

[0016] Optionally, it also includes a method for performing a blowing operation, the method comprising: Step S420: When there are bud and leaf features but no vegetation branches or trunks, obtain the current environmental wind speed; Step S421: If the current ambient wind speed falls within the preset effective wind speed range, determine the initial bud position based on the characteristics of the bud and branches, and obtain the current bud position in real time to determine the change in bud position. Step S422: Determine the area occupied by buds based on the characteristics of buds and branches, and determine the range of changes in the location of abnormal buds based on the area occupied by buds; Step S423: If the change in bud position falls within the range of abnormal bud position changes, output the empty area signal; Step S424: If the change in bud position does not fall within the range of abnormal bud position changes, no empty area signal is output. Step S425: If the current ambient wind speed does not fall within the effective wind speed range, determine the blowing position and blowing force based on the initial bud point position, and form a blowing plan; Step S426: Control the preset vacancy detection device to perform the blowing operation according to the blowing scheme, and obtain the current bud position in real time to determine the amount of change in bud position, so as to execute steps S422 to S424.

[0017] By employing the above technical solution, the change in bud position is calculated by comparing the initial bud position with the current position. Simultaneously, the area occupied by the bud and its branches is determined based on their characteristics, and the range of abnormal bud position changes is further determined based on this area. If the calculated change in bud position falls within this abnormal range, it indicates that the bud and its branches may not be a valid plant rooted in the planting substrate, but rather scattered debris carried by the wind; in this case, an empty area signal is output. Conversely, if the change in bud position does not fall within the abnormal range, it indicates that the bud and its branches are relatively stable and may be part of a valid plant; in this case, no empty area signal is output.

[0018] Optionally, if the change in bud position falls within the range of abnormal bud position changes, the method for outputting the vacancy region signal includes: Step S4230: Obtain the image information of the initial position of the bud corresponding to the initial bud position; Step S4231: Determine the original bud features based on the initial position image information of the buds, and output them as bud and leaf features; Step S4232: If the original bud point feature exists, repeat steps S420 to S425.

[0019] By adopting the above technical solution, when the change in bud position falls within the abnormal range, the initial bud position image information is obtained, and the original bud features that are obscured by surface buds are identified. If the original bud features exist, secondary wind force detection is performed on the original true buds. This avoids misjudgments caused by surface floating buds obscuring true buds, performs secondary verification on buds with abnormal displacement, improves the accuracy of bud detection, and reduces the output of invalid vacancy signals.

[0020] Optionally, if no growth transition connection region exists, methods for outputting vacancy region signals include: Step S440: Determine the current number of branches and leaves and the current orientation of branches and leaves based on the characteristics of buds and branches; Step S441: If the current number of branches and leaves is not 1, determine the current branch orientation based on the characteristics of the vegetation branches; Step S4410: If the current number of branches and leaves is 1, output the empty area signal; Step S442: Determine the consistency quantity based on the current branch and leaf orientation and the current branch orientation; Step S443: If the number of consistent elements is equal to the current number of branches and leaves, do not output the empty area signal; Step S444: If the number of consistent vegetation is less than the current number of branches and leaves, determine the location of suspected abnormal vegetation; Step S445: Determine the blowing location and blowing force based on the suspected abnormal vegetation location to form a blowing plan, and execute the blowing operation according to the blowing plan; Step S446: Obtain the number of fallen branches and leaves during the blowing operation; Step S447: If the number of fallen branches and leaves is equal to the current number of branches and leaves, output the empty area signal.

[0021] By adopting the above technical solution, when it is impossible to determine the effective vegetation through the growth transition connection area, the current orientation of the branches and leaves and the current orientation of the branches and trunks can be used to determine whether the branches and leaves and the branches and trunks are growing as one, so as to avoid misjudging the overlapping debris as effective vegetation and improve the reliability of the gap detection.

[0022] Optionally, methods for performing the blowing operation include: Step S448: Obtain the location of the vacancy detection device and determine the current vacancy detection device number based on the air blowing position; Step S449: Determine the sliding amount, lifting and lowering amount, yaw angle adjustment amount, and pitch angle adjustment amount of the moving support based on the current empty space detection device number to form the empty space detection device movement parameters; Step S450: When there are no empty space detection device movement parameters, determine the current planting trough number based on the air blowing position; Step S451: Determine the valid empty space detection device number corresponding to the current planting trough number based on the current planting trough number, and re-determine the empty space detection device movement parameters; Step S452: Based on the movement parameters of the empty space detection device and the blowing force, a blowing plan is formed, and the blowing operation is performed according to the blowing plan.

[0023] By adopting the above technical solution, the device number is matched according to the principle of proximity, and the movement and angle adjustment parameters are calculated. When no parameters are available, the planting trough number is determined by the air blowing position, and the corresponding dedicated effective device is matched, and the parameters are recalculated. The parameters and air blowing force are combined to form a plan, and the air blowing operation is executed. This provides a means for the nearest empty space detection device to perform the air blowing operation, and solves the problem of the inaccessibility of the nearest empty space detection device, ensuring the feasibility of the air blowing verification process and supporting the smooth completion of empty space detection.

[0024] Secondly, the present invention provides a plant status and vacancy detection system based on image recognition, which adopts the following technical solution: A plant status and vacancy detection system based on image recognition, comprising: The acquisition module is used to acquire image information of the planting area; The memory is used to store the program of the image recognition-based plant status and vacancy detection method described above; The processor loads and executes programs from memory.

[0025] By adopting the above technical solution, the acquisition module can collect images of the planting area, providing raw data support for subsequent plant status analysis and vacancy detection; the memory is responsible for storing the program code that implements the detection method, ensuring the stable operation of the method and the reliable preservation of data; the processor, as the core of the system, loads and executes the program in the memory to perform a series of complex processing and analysis on the image information collected by the acquisition module, including vegetation feature extraction, growth trend judgment, and vacancy area identification, ultimately achieving accurate assessment of plant status and vacancy detection.

[0026] In summary, the present invention has at least one of the following beneficial technical effects: By using identification logic such as soil-adhering branch recognition, growth transition connection area judgment, and wind displacement verification, it can distinguish between dead branches and leaves, overlapping branches and leaves and other debris and real living plants, reducing the situation where traditional image recognition easily misjudges debris as vegetation. By combining the growth trends of adjacent vegetation and the statistical patterns of gap size, it can automatically identify reasonable gaps reserved for vegetation expansion, avoid misjudging normal growth gaps as planting deficiencies, and make the detection results more in line with actual planting planning needs. When the natural wind speed does not meet the testing conditions, active air blowing verification is initiated. The empty space testing device is scheduled according to the rules of proximity priority and planting trough exclusive binding to ensure the smooth execution of the air blowing verification process. Attached Figure Description

[0027] Figure 1 This is a flowchart of a plant status and vacancy detection method based on image recognition in an embodiment of this application; Figure 2 This is a schematic diagram of the vacancy detection device in the embodiments of this application; Figure 3 This is an exploded view of the pitch angle adjustment component and the yaw angle adjustment component in the embodiments of this application; Figure 4 This is an exploded view of the lifting block and yaw angle adjustment component in the embodiments of this application.

[0028] The parts referred to by the numbers in the above attached diagrams are as follows: 1. Greenhouse support frame; 2. Movable support frame; 21. Horizontal sliding groove; 3. Lifting block; 31. Lifting groove; 4. Yaw angle adjustment assembly; 41. Yaw angle adjustment seat; 42. First yaw angle adjustment lug; 43. Yaw angle adjustment rotating shaft; 44. Second yaw angle adjustment lug; 5. Pitch angle adjustment assembly; 51. Pitch angle adjustment base; 52. First pitch angle adjustment lug; 53. Pitch angle adjustment rotating shaft; 54. Second pitch angle adjustment lug; 6. Blower; 7. Camera. Detailed Implementation

[0029] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments.

[0030] This invention discloses a method for plant status and void detection based on image recognition. (Refer to...) Figure 1 A method for plant status and vacancy detection based on image recognition includes: Step S1: Obtain image information of the planting area.

[0031] The image information of the planting area refers to the digital image of the entire planting trough area obtained by sequentially capturing images of individual planting areas through image acquisition devices installed at designated monitoring locations within the planting area, followed by image integration processing. This image includes visual pixel information such as the planting substrate, living vegetation, empty areas, and dead branches and debris. The overall digital image of the planting trough area is obtained by integrating the images of the individual planting areas captured. A planting trough is a long, narrow, or modular cultivation carrier within the planting area used to support the planting substrate and arranged according to a plan.

[0032] Because plants may wither or be damaged during the planting process, some areas may not be planted properly, or withered plants may leave empty spaces. Therefore, it is necessary to monitor the plant growth status in real time to identify and determine the empty areas.

[0033] Step S2: Analyze the image information of the planting area to obtain the current vegetation features.

[0034] Current vegetation features refer to the visual features of living vegetation extracted from the image information of the planting area, including leaf color, outline, leaf vein texture, plant morphology, stem features, and spatial distribution features.

[0035] Step S20: If current vegetation features exist, determine the current vegetation category based on the current vegetation features.

[0036] The current vegetation category refers to the type of artificially cultivated vegetation variety identified based on vegetation characteristics, such as leafy vegetables, fruits and vegetables, herbaceous plants, and other preset vegetation types.

[0037] If the current vegetation features are present, it means that there are plants with the visual characteristics of living vegetation in the area corresponding to the planting trough, and it is not a purely empty space or a cluttered area.

[0038] Step S21: Determine the current vegetation growth status based on the current vegetation category and current vegetation characteristics, and record it in the preset vegetation information table.

[0039] The current vegetation growth status refers to the plant growth condition determined by the vegetation category growth standards, and is divided into normal vegetation growth status, mature vegetation growth status, and abnormal vegetation growth status. Abnormal vegetation status is specifically manifested by yellowing, wilting, and non-living dead branches and debris.

[0040] A vegetation information table is a structured record table established based on the vegetation covered by the planting area, used to store data such as vegetation type, growth status and monitoring time for each planting area.

[0041] Step S22: If the current vegetation growth state is abnormal, obtain the coordinates of the abnormal vegetation growth area and output them.

[0042] The coordinates of the abnormal vegetation growth area refer to the physical location coordinates of the abnormally growing vegetation in the actual planting area. Specifically, they need to be obtained from the placement of camera 7 and the shooting angle corresponding to the image information of the planting area.

[0043] If the current vegetation growth status is abnormal, it indicates that the living vegetation in the area is experiencing poor growth, withering and decay, or being disturbed by non-living debris, and a marker should be placed to indicate this.

[0044] Step S23: When the current vegetation category does not exist, identify the abnormal vegetation area.

[0045] Abnormal vegetation areas refer to planting areas that have visual characteristics of branches and leaves that resemble vegetation, but cannot be matched with any vegetation category. These areas are mostly areas with dead branches and fallen leaves, scattered branches, and interference from non-cultivated plants.

[0046] If the current vegetation category does not exist, it means that the area only has the appearance of vegetation and is not the target cultivated living plant. Further identification is needed to determine whether it is pseudo-vegetation interference.

[0047] Step S24: Obtain image information of abnormal vegetation areas based on abnormal vegetation areas, and analyze the image information of abnormal vegetation areas to determine the characteristics of vegetation branches and trunks.

[0048] Abnormal vegetation area image information refers to image data obtained by re-photographing abnormal vegetation areas and then performing noise reduction and enhancement.

[0049] Vegetation branch characteristics refer to the characteristic information extracted from abnormal vegetation areas, such as branch outline, extension posture, surface texture, and contact pattern with planting substrate.

[0050] Step S25: Determine the soil-adhering branches based on the characteristics of vegetation branches.

[0051] "Soil-adherent branches" refers to a branch morphology within abnormal vegetation areas where branches lie flat and adhere to the planting substrate surface without any upward growth, and the branches are seamlessly attached to the substrate surface. Image recognition technology is used to identify the ends of the branches furthest from and closest to the planting substrate, and to observe whether both ends are in contact with the substrate surface. The planting substrate refers to the soil used to anchor plant roots, provide nutrients, and offer support.

[0052] Step S26: When there is soil adhering to the branches, determine the abnormal growth and rooting characteristics.

[0053] When the soil is attached to the branch, it indicates that the branch is a fallen, flat, dead branch, not a living plant branch that has taken root and grown from the planting substrate.

[0054] Step S27: When abnormal growth and rooting characteristics exist, output the vacancy region signal.

[0055] Abnormal growth and rooting characteristics refer to branches that merely lie flat against the surface of the planting substrate without growing upwards from the inside of the substrate, which is a typical characteristic for identifying non-living dead branches.

[0056] Vacancy zone signal refers to a status marker signal used to indicate that there is no effective living vegetation in the corresponding area of ​​the planting trough and it is judged as an abnormal vacancy.

[0057] When abnormal growth and rooting characteristics are present, it indicates that the area only contains non-living dead branches and debris, and no living cultivated plants, and should be identified as an empty area.

[0058] Step S3: When there are no current vegetation features, output the vacancy area signal.

[0059] If the current vegetation features are not present, it means that the area corresponding to the planting trough has neither living plants nor dead branches and debris, and is a pure empty area without vegetation cover.

[0060] This also includes: Step S40: When there are no vegetation branch features or no abnormal growth and rooting features, obtain bud and leaf features based on the image information of the abnormal vegetation area.

[0061] Bud and leaf features refer to the morphological, color, and texture characteristics of living growth tissues such as buds, new leaves, and young branches identified from images of abnormal vegetation areas.

[0062] When there are no vegetation branches, it means that there are no branches in the area. The vegetation may be in the budding stage, and the features obtained at this stage cannot meet the judgment of the current vegetation category. Therefore, it is necessary to obtain the bud and leaf features for analysis. When there are no abnormal growth and rooting features, it means that there are branches in the area. The branches are not fallen or dead branches. There may be branches that are vertically inserted into the soil.

[0063] Step S41: When there are no buds or branches, output the empty area signal.

[0064] When there are no buds or leaves, it indicates that there are two situations in the area. The first is that there are neither effective branches nor new buds and leaves, and no signs of living vegetation. This should be judged as an empty space, so the empty space area signal is directly output. The second is that there are effective branches, but no leaves attached to the branches. In this embodiment, there is no vegetation category where any plant has only branches but no leaves during its growth stage, so the empty space area signal is directly output.

[0065] Step S42: When there are bud and leaf characteristics and vegetation branch and trunk characteristics, determine the growth transition connection area based on the bud and leaf characteristics and vegetation branch and trunk characteristics.

[0066] The growth transition connection zone refers to the connection area formed by the natural connection and continuous growth between buds, leaves and branches. It is characterized by the seamless integration of the ends of leaves and branches with the surface of branches. Specifically, the leaves and branches are naturally fused with the surface of branches through vascular bundles, without any trace of breakage, splicing or overlapping.

[0067] When bud and leaf characteristics are present, it indicates the presence of buds and leaves resembling living vegetation in the area. However, it is still necessary to determine whether they are naturally grown from the same plant as the branches and trunk. This is because there may be interference from leaves naturally falling onto the branches and trunk, causing the bud and leaf characteristics to not grow as a single entity from the branches and trunk.

[0068] Step S43: If there is no growth transition connection region, output the vacancy region signal.

[0069] When there is no growth transition connection area, it means that the buds and branches are just scattered pieces spliced ​​together and not naturally grown as a whole. They are mostly fallen branches and leaves, not real plants.

[0070] Step S44: If a growth transition connection region exists, output an abnormal vegetation region signal.

[0071] Abnormal vegetation area signal refers to a status marker signal used to indicate that there are living plants in the corresponding area of ​​the planting trough, but the vegetation type cannot be identified, indicating that manual intervention is required for verification.

[0072] When a transitional growth zone exists, it indicates that there are indeed living plants in that area, but the specific vegetation type cannot be identified. This is considered an abnormal vegetation area, and manual confirmation is required.

[0073] This also includes a method for outputting a reasonable empty space region signal, which includes: Step S45: When there is a vacancy signal, search for vegetation characteristics in the adjacent area.

[0074] Adjacent vegetation features refer to the visual characteristics of adjacent vegetation within a preset detection range around the empty area to be detected, such as crown size, branch and leaf extension direction, and plant growth. The detection range is a rectangular detection area formed by extending a preset detection pixel distance to the left and right of the empty area to be detected, used to obtain reference information about the surrounding vegetation. The detection pixel distance is a pre-set pixel length value used to limit the size of the detection range for adjacent vegetation features. This pixel length can be converted into an actual detection distance in physical space through the coordinate calibration relationship between the image and the actual planting area.

[0075] When a signal for an empty space is present, it indicates that the current area has been preliminarily identified as an empty space, and further assessment is needed to determine whether the empty space is a reasonably reserved space. A reasonably reserved space refers to planting space planned in advance for the subsequent growth and expansion of vegetation in the surrounding area that exhibits lateral expansion and canopy extension characteristics. It is not an abnormal empty space formed due to missed planting, withered plants, or missing plants.

[0076] Step S46: Determine the vegetation growth trend of adjacent areas based on the vegetation characteristics of adjacent areas.

[0077] The vegetation growth trend in adjacent areas refers to the direction of future spatial growth and expansion of vegetation determined based on the current crown width, branch and leaf extension direction and growth characteristics of adjacent vegetation.

[0078] Step S47: If the vegetation growth trend in the adjacent area is an expansion trend, output a reasonable vacancy area signal.

[0079] The tendency of peripheral expansion growth refers to the growth characteristic of adjacent plants that extend laterally in all directions and gradually expand their crown width during growth, requiring additional space.

[0080] The reasonable vacancy area signal refers to the reasonable space that is pre-planned and reserved to accommodate the subsequent expansion and growth of vegetation in the vacancy area, and is not a marker signal for abnormally missing vacancy areas.

[0081] When the vegetation growth trend in the adjacent area is an expansion trend, it indicates that the empty area is a reasonable space reserved to accommodate the subsequent expansion of vegetation, and a reasonable empty area signal should be output.

[0082] Step S48: If the vegetation growth trend in the adjacent area is not the same as the surrounding expansion growth trend, output the vacant area signal.

[0083] When the vegetation growth trend in the adjacent area does not follow the surrounding expansion trend, it indicates that there is no need to reserve space for vegetation growth in the empty area, which is an abnormal empty space caused by planting errors or missing plants.

[0084] Step S49: When there is no vegetation growth trend in adjacent areas, determine the adjacent vegetation category based on the vegetation characteristics of adjacent areas.

[0085] Adjacent vegetation category refers to the classification of vegetation species or growth type based on the vegetation characteristics of adjacent areas.

[0086] When there is no vegetation growth trend in adjacent areas, the vegetation may be in the early stage of growth. The growth trend cannot be directly judged from the current image features. It is necessary to further query its inherent growth habits through vegetation category.

[0087] Step S50: Search the preset vegetation feature database based on the adjacent vegetation categories to determine the growth trend of adjacent vegetation, and then proceed to steps S47 to S48.

[0088] A vegetation characteristic database refers to a database that is pre-built and stores standard characteristic information such as the growth habits, expansion trends, and spatial requirements of various types of vegetation that need to be observed in greenhouses.

[0089] The growth trend of adjacent vegetation refers to the inherent growth and expansion characteristics of vegetation corresponding to adjacent vegetation categories.

[0090] Steps S47 to S48 are re-executed based on the growth trend of adjacent vegetation to output a reasonable vacancy area signal or a vacancy area signal.

[0091] This also includes a method for outputting a reasonable vacancy area signal when there are no adjacent vegetation types, the method comprising: Step S450: Obtain the current example region.

[0092] The current example area refers to the standardized statistical area defined with the target empty space area as the center, according to the preset example detection range.

[0093] The example detection range refers to the pre-defined physical space range centered on the target empty area and extending along the planting trough to cover several consecutive planting units before and after it.

[0094] Step S451: Determine the empty space area and the number of empty space areas based on the current example area.

[0095] The number of empty areas refers to the total number of empty areas identified and determined within the current example area.

[0096] Step S452: Determine the length of the empty space area based on the empty space area.

[0097] The length of the empty space refers to the actual physical length of a single empty space along the direction of the planting trough.

[0098] Step S453: Determine the total length of vacant spaces based on the length of the vacant areas, and determine the average length of vacant spaces based on the number of vacant areas.

[0099] The average length of vacant space occupancy refers to the ratio of the total length of all vacant spaces occupies in the current example area to the number of vacant spaces, which is used to reflect the typical size of vacant spaces in the area.

[0100] Step S454: Determine the deviation value of the occupied length based on the occupied length of the vacant area and the average occupied length of the vacant area.

[0101] The deviation value of the occupied length refers to the absolute value of the difference between the occupied length of a single vacant area and the average occupied length of vacant areas. It can be used to measure whether the length of the vacant area conforms to the overall distribution pattern of the area.

[0102] Step S455: If the deviation value of the occupied length falls within the preset reasonable deviation value range, output a reasonable empty space area signal.

[0103] The reasonable deviation range refers to the pre-set length deviation interval used to determine whether the length of the empty space conforms to the manually planned layout.

[0104] When the deviation value of the occupied length falls within the reasonable deviation value range, it indicates that the occupied length of the vacant area is consistent with the overall vacant distribution pattern in the area. It belongs to the reasonable vacant area reserved by unified artificial planning, rather than the abnormal vacant area formed by random absence. Therefore, the reasonable vacant area signal is output.

[0105] This also includes a method for outputting a reasonable vacancy area signal if the deviation value of the occupied length does not fall within the reasonable deviation value range. This method includes: Step S4550: Determine the current empty space region based on the current example region.

[0106] The current empty space area refers to the target empty space area within the current example area that has been determined to have a length deviation value that exceeds the reasonable deviation value range, and requires further verification of the arrangement rules.

[0107] Step S4551: Obtain the left and right adjacent vegetation areas corresponding to the current empty space area.

[0108] The adjacent vegetation area to the left refers to the vegetation distribution area that is immediately to the left of the current empty space along the direction of the planting trough.

[0109] The adjacent vegetation area to the right refers to the vegetation distribution area immediately to the right of the current empty space along the direction of the planting trough.

[0110] Step S4552: Obtain the left adjacent region corresponding to the left adjacent vegetation region.

[0111] The left adjacent area refers to the vegetation distribution area immediately to the left of the left adjacent vegetation area along the extension direction of the planting trough.

[0112] Step S4553: Obtain the right adjacent region corresponding to the right adjacent vegetation region.

[0113] The right adjacent area refers to the vegetation distribution area immediately to the right of the right adjacent vegetation area along the extension direction of the planting trough.

[0114] Step S4554: If both the left and right adjacent areas are empty areas, output a reasonable empty area signal.

[0115] When both the left and right adjacent areas are empty, it means that the areas on both sides of the left and right adjacent vegetation areas are empty. In this case, the current empty area conforms to the pattern of artificial planting layout and belongs to a reasonable empty space reserved for vegetation growth. Therefore, a reasonable empty area signal is output.

[0116] This also includes a method for performing a blowing operation, the method comprising: Step S420: When there are bud and leaf features but no vegetation branches or trunks, obtain the current environmental wind speed.

[0117] Current ambient wind speed refers to the airflow speed data collected in real time within the planting area by a wind speed sensor. When there are bud and leaf characteristics but no vegetation trunk characteristics, the small size and light weight of pseudo-buds, new leaves, and other debris, which are devoid of vital signs, make them easily affected by ambient airflow and cause them to sway. The wind speed can be used to determine whether they are actually living tissues attached to the surface of the planting substrate.

[0118] When there are bud and leaf characteristics but no vegetation trunk characteristics, it means that the area only has bud and leaf vegetation characteristics and no trunk structure. It is impossible to determine the authenticity by the connection of trunk and branch. Natural wind is needed to help determine whether the buds and leaves are rooted and alive.

[0119] Step S421: If the current ambient wind speed falls within the preset effective wind speed range, determine the initial bud position based on the characteristics of the bud and branches, and obtain the current bud position in real time to determine the change in bud position.

[0120] The effective wind speed range refers to the range of natural wind speeds determined through multiple experiments that can blow loose fallen leaves without shaking rooted seedlings. Wind speeds that are too low will not have a testing effect, while wind speeds that are too high will easily affect normal plants.

[0121] The initial bud position refers to the baseline coordinate position of the bud and its branches in the first image acquired before wind speed detection. The change in bud position refers to the coordinate offset difference between the real-time bud position and the initial bud position, which reflects the degree of movement of the bud due to wind force.

[0122] The current bud location refers to the actual coordinate position of the bud and its branches in the real-time acquired image during the effective wind speed range.

[0123] If the current ambient wind speed falls within the effective wind speed range, it means that the current natural wind force is suitable as a detection condition, and the bud displacement status can be used to determine whether it is a living organism with fixed roots.

[0124] Step S422: Determine the area occupied by the buds based on the characteristics of the buds and branches, and determine the range of changes in the location of abnormal buds based on the area occupied by the buds.

[0125] The area occupied by a bud refers to the actual size of the area covered by the projection of the bud's branches and leaves onto the surface of the planting substrate.

[0126] The range of abnormal bud position changes refers to the displacement threshold interval determined by referring to a preset mapping table of bud size and position changes based on the bud's own dimensions. Falling into this interval indicates that the bud has significantly shifted due to wind force and is in a non-fixed state. The mapping table of bud size and position changes is a table relating the range of abnormal bud position changes to the area occupied by the bud, obtained from multiple experiments. Specifically, if the bud occupies an area of ​​x cm in length and y cm in width, then the range of abnormal bud position changes is the interval between a movement of x cm along the length of the bud's area and a movement of y cm along the width of the bud's area. This is because when the bud moves x cm along the length of its area or y cm along the width of its area, it indicates that the bud has moved from its initial position, thus confirming that it is not rooted in the soil.

[0127] Step S423: If the change in bud position falls within the range of abnormal bud position changes, output the empty area signal.

[0128] When the change in bud position falls within the range of abnormal bud position changes, it indicates that the bud has been significantly shifted by natural wind. The bud is not fixed to the planting substrate and is only a scattered piece of debris that has fallen, without any living vegetation characteristics, and is therefore identified as an empty area.

[0129] Step S424: If the change in bud position does not fall within the range of abnormal bud position changes, no empty area signal is output.

[0130] When the change in bud position does not fall within the range of abnormal bud position changes, it indicates that the bud is only slightly swayed by the wind and has no obvious displacement. The base is fixed to the planting substrate, and it belongs to a living seedling that has taken root and is not identified as an empty area.

[0131] Step S425: If the current ambient wind speed does not fall within the effective wind speed range, determine the blowing position and blowing force based on the initial bud point position, and form a blowing plan.

[0132] The airflow position refers to the airflow range determined by a preset airflow distance offset from the edge of the planting trough towards the center, targeting the area where the buds and leaves are located. This distance is dynamically adjusted according to the size of the bud's area to ensure that the airflow effectively reaches the base of the bud. The airflow distance is a value positively correlated with the bud's area, obtained through multiple airflow tests; that is, the larger the bud's area, the closer the airflow distance, ensuring that the airflow accurately reaches the root area of ​​the bud. Due to the limited space for the empty space detection device to move, refer to... Figure 2 For example, in the horizontal direction, it can only slide along the length of the horizontal sliding groove 21, so it only needs to fall into the air flow range corresponding to the blowing position to perform the blowing operation.

[0133] The wind power refers to the output power of the equipment that simulates natural wind. Its magnitude must be matched with the wind power value corresponding to the effective wind speed range to ensure that it can blow non-fixed buds and leaves without damaging the rooted seedlings.

[0134] A blowing scheme refers to a standardized airflow action plan formed by combining blowing location, blowing force, and preset blowing duration parameters. The blowing duration parameter refers to the duration of airflow action set according to the area occupied by the bud. The larger the area occupied, the longer the blowing duration, to ensure that non-fixed buds can fully respond to the wind force and generate displacement.

[0135] If the current ambient wind speed is not within the effective wind speed range, it indicates that the natural wind conditions are insufficient or too strong, and cannot be directly used to detect whether the buds have taken root. In this case, it is necessary to use the blower device 6 to actively blow air in order to detect changes in the displacement of the buds.

[0136] Step S426: Control the preset vacancy detection device to perform the blowing operation according to the blowing scheme, and obtain the current bud position in real time to determine the amount of change in bud position, so as to execute steps S422 to S424.

[0137] This embodiment discloses a vacancy detection device, referring to... Figure 2 The vacancy detection device is installed at one end of the greenhouse support 1 near the planting trough. The vacancy detection device includes a movable support 2, a lifting block 3, a yaw angle adjustment component 4, a pitch angle adjustment component 5, a blower 6, and a camera 7.

[0138] Reference Figure 3 and Figure 4 One end of the movable support 2 is fixedly connected to the end of the greenhouse support 1 near the planting trough. Horizontal sliding grooves 21 are provided on both sides of the movable support 2. The lifting block 3 is embedded in the horizontal sliding groove 21 and slides along the length of the horizontal sliding groove 21. A lifting groove 31 is provided at the end of the lifting block 3 away from the movable support 2.

[0139] Reference Figure 3 The yaw angle adjustment assembly 4 includes a yaw angle adjustment seat 41, a first yaw angle adjustment lug 42, a yaw angle adjustment rotating shaft 43, and a second yaw angle adjustment lug 44. One end of the yaw angle adjustment seat 41 is embedded in the lifting groove 31 and slides along the length of the lifting groove 31. One end of the first yaw angle adjustment lug 42 is fixedly connected to the end of the yaw angle adjustment seat 41 away from the lifting groove 31. The yaw angle adjustment rotating shaft 43 is rotatably connected to the first yaw angle adjustment lug 42. The yaw angle adjustment rotating shaft 43 is fixedly connected to the second yaw angle adjustment lug 44.

[0140] Reference Figure 3The pitch angle adjustment assembly 5 includes a pitch angle adjustment base 51, a first pitch angle adjustment lug 52, a pitch angle adjustment rotating shaft 53, and a second pitch angle adjustment lug 54. One end of the pitch angle adjustment base 51 is fixedly connected to the end of the second pitch angle adjustment lug 54 away from the lifting block 3. One end of the first pitch angle adjustment lug 52 is fixedly connected to the end of the pitch angle adjustment base 51 away from the second pitch angle adjustment lug 44. The pitch angle adjustment rotating shaft 53 is rotatably connected to the first pitch angle adjustment lug 52. The second pitch angle adjustment lug 54 and the pitch angle adjustment rotating shaft 53 are fixedly connected. The end of the second pitch angle adjustment lug 54 away from the lifting block 3 is fixedly connected to the blower 6. The camera 7 is fixedly connected to one side of the second pitch angle adjustment lug 54.

[0141] Among them, if the change in bud position falls within the range of abnormal bud position changes, the methods for outputting the vacancy region signal include: Step S4230: Obtain the image information of the initial position of the bud corresponding to the initial bud position.

[0142] The initial location image information of the bud points refers to the baseline image information collected before the action of natural wind speed, which includes the area covered by the initial bud points and the planting substrate below it. It is used to record the initial coverage state of the bud points and the hidden features of the underlying layer.

[0143] Step S4231: Determine the original bud features based on the initial position image information of the buds, and output them as bud and leaf features.

[0144] The original bud characteristics refer to the morphology, color, and growth attachment characteristics of the actual buds that exist at the bottom layer and are obscured by the surface floating buds and leaves, thus failing to be identified initially. For example, fallen leaves may cover buds, causing them to go unrecognized at first. Specifically, the original bud characteristics here refer to the actual buds that were originally covered by the surface floating buds after they were blown away by the wind.

[0145] Step S4232: If the original bud point feature exists, repeat steps S420 to S425.

[0146] When the original bud features are present, it means that the previously determined abnormal displacement was only the floating buds or fallen leaves scattered on the surface. After they were blown away by the wind, the real rooted buds that were originally covered were exposed. It is necessary to re-verify the area and displacement changes of the original real buds. Therefore, steps S420 to S425 are executed.

[0147] If no growth transition connection region exists, the methods for outputting the vacancy region signal include: Step S440: Determine the current number of branches and leaves and the current orientation of branches and leaves based on the characteristics of buds and branches.

[0148] The current number of branches and leaves refers to the total number of independent buds and young leaves identified from the image of abnormal vegetation area, which is the number of individual branches and leaves after excluding the branch structure.

[0149] The current branch and leaf orientation refers to the single branch and leaf growth vector determined by the branch and leaf extension direction and the leaf vein extension direction extracted for each individual branch and leaf, which represents the natural growth posture of the branch and leaf.

[0150] Step S441: If the current number of branches and leaves is not 1, determine the current branch orientation based on the characteristics of vegetation branches.

[0151] The current branch orientation refers to the direction of the main branch and its extension extracted from the image of the abnormal vegetation area. This direction of extension represents the overall growth direction of the branch.

[0152] If the current number of branches and leaves is not 1, it means that there are different branches and leaves shading each other, making it impossible to identify the growth transition connection area through image recognition. Therefore, further analysis is needed based on the current branch orientation and the current branch and leaf orientation.

[0153] Step S4410: If the current number of branches and leaves is 1, output the empty area signal.

[0154] Reference Figure 2 The number of empty space detection devices is set to several groups, and the number of groups is determined according to the assembly space of the greenhouse support 1 and the number of movable supports 2. When the growth transition area between the branch and the leaves attached to its surface cannot be identified, the cameras of other empty space detection devices will be controlled to take pictures, that is, the shooting angle will be updated. Therefore, when the current number of leaves is 1, it means that there is no situation where different leaves block each other on the branch. In the case that the current vegetation type cannot be identified and the branches and leaves have no vital signs, the empty space area signal can be directly output.

[0155] Step S442: Determine the consistency quantity based on the current branch and leaf orientation and the current branch orientation.

[0156] The consistency count refers to the number of branches and leaves whose growth direction matches the branch's direction within a preset growth direction matching threshold. In other words, it's the number of branches and leaves whose growth direction is consistent with the branch's growth direction. The growth direction matching threshold is an angular range obtained through multiple experiments to determine whether the growth direction of branches and leaves is coordinated with the extension direction of the branch. The branch may have multiple branches, and branches and leaves located on different branches are judged according to the extension direction of their corresponding branches using the growth direction matching threshold. Furthermore, because the growth direction matching threshold varies for different vegetation types, the consistency count alone cannot directly determine whether the vegetation in a region is pseudo-vegetation. Therefore, subsequent verification will incorporate wind blowing operations.

[0157] Step S443: If the number of consistent elements is equal to the current number of branches and leaves, do not output the empty area signal.

[0158] When the number of consistent branches equals the current number of branches and leaves, it indicates that the growth direction of all branches and leaves in this area is highly consistent with the direction of the branch trunk, conforming to the natural growth polarity of the same living plant. Although no growth transition connection area is identified, it is still determined to be a real and valid plant, and no empty area signal is output. There are two situations where no growth transition connection area is identified: the first is pseudo-vegetation formed by branches and fallen leaves without any living characteristics; the second is the situation where branches and leaves block each other, that is, branches and leaves A block the growth transition area between branches and leaves and their trunk.

[0159] Step S444: If the number of consistent vegetation is less than the current number of branches and leaves, determine the location of suspected abnormal vegetation.

[0160] The suspected abnormal vegetation location refers to the location of the branches and trunks corresponding to branches and leaves whose growth direction does not fall within the growth direction matching threshold, and also includes the comprehensive coordinate range of the area where the branches and leaves with mismatched growth direction are located. In other words, it is the entire area of ​​branches and trunks suspected of having scattered branches and leaves with overlapping debris.

[0161] When the number of consistent branches is less than the current number of branches and leaves, it indicates that the growth direction of some branches and leaves in this area does not match the growth direction of the trunk, and the unity of branches and leaves with the trunk cannot be proved by the synergy of growth direction.

[0162] Step S445: Determine the blowing location and blowing force based on the suspected abnormal vegetation location to form a blowing plan, and execute the blowing operation according to the blowing plan.

[0163] Step S446: Obtain the number of fallen branches and leaves during the blowing operation.

[0164] The number of fallen branches and leaves refers to the total number of branches and leaves that have obviously shifted from the suspected abnormal vegetation location, detached from the corresponding branches and planting substrate, and fallen during the wind blowing operation.

[0165] Step S447: If the number of fallen branches and leaves is equal to the current number of branches and leaves, output the empty area signal.

[0166] When the number of fallen leaves equals the current number of leaves, it indicates that all leaves in the area are scattered debris attached to the branches and trunks, and have not formed a naturally growing integrated connection with the branches and trunks. They have completely fallen off under the action of wind, and there is no effective living vegetation in the area. Therefore, an empty area signal is output.

[0167] The methods for performing the blowing operation include: Step S448: Obtain the location of the vacancy detection device and determine the current vacancy detection device number based on the air blowing position.

[0168] The location of the vacancy detection device refers to the real-time coordinates of each vacancy detection device on the greenhouse support 1, determined by the scale of the horizontal sliding groove 21 of the moving support 2, the height parameters of the lifting groove 31, and the horizontal sliding area corresponding to the current vacancy detection device number. (Refer to...) Figure 2 The lower edge of the greenhouse support 1 is equipped with multiple sets of empty space detection devices in the planting trough array. Each set of empty space detection devices can only slide along the length direction of its horizontal sliding groove 21.

[0169] The current vacancy detection device number refers to the unique identifier of the vacancy detection device that is closest to the blowing position, selected based on the target coordinates of the blowing position according to the principle of proximity.

[0170] Step S449: Determine the sliding amount, lifting and lowering amount, yaw angle adjustment amount, and pitch angle adjustment amount of the moving support based on the current empty space detection device number to form the empty space detection device movement parameters.

[0171] The sliding distance of the movable support refers to the horizontal movement distance required for the lifting block 3 to move along the horizontal sliding groove 21 of the movable support 2.

[0172] The lifting and sliding amount refers to the vertical movement distance required for the yaw angle adjustment seat 41 along the lifting groove 31 of the lifting block 3, which is used to adjust the height of the empty position detection device to match the blowing position.

[0173] The yaw angle adjustment amount refers to the horizontal rotation angle required for the yaw angle adjustment rotating shaft 43, which is used to adjust the horizontal air delivery direction of the blower 6.

[0174] The pitch angle adjustment amount refers to the pitch rotation angle required for the pitch angle adjustment rotating shaft 53, which is used to adjust the up and down air supply pitch angle of the blower 6.

[0175] The movement parameters of the vacancy detection device refer to the driving parameters formed by combining the above-mentioned sliding amount and angle adjustment amount.

[0176] Step S450: When there are no empty space detection device movement parameters, determine the current planting trough number based on the air blowing position.

[0177] When there are no movement parameters for the empty space detection device, it means that the nearest matching empty space detection device is limited by the structure of horizontal sliding and lifting stroke, and cannot reach the blowing position through multi-degree-of-freedom adjustment.

[0178] Step S451: Determine the valid empty space detection device number corresponding to the current planting trough number based on the current planting trough number, and re-determine the empty space detection device movement parameters.

[0179] The current planting trough number refers to the unique identifier number of the planting trough corresponding to the air blowing position.

[0180] The effective empty space detection device number refers to the empty space detection device number that is pre-bound to each planting trough number. The travel range of this device completely covers all detection areas of the corresponding planting trough, and it has the ability to detect the entire area of ​​the planting trough.

[0181] Step S452: Based on the movement parameters of the empty space detection device and the blowing force, a blowing plan is formed, and the blowing operation is performed according to the blowing plan.

[0182] Based on the same inventive concept, embodiments of the present invention provide a plant status and vacancy detection system based on image recognition.

[0183] A plant status and vacancy detection system based on image recognition, comprising: The acquisition module is used to acquire image information of the planting area; The memory is used to store a program for a plant status and vacancy detection method based on image recognition; The processor loads and executes programs from memory.

[0184] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. An image recognition-based plant status and space detection method, characterized by, include: Step S1: Obtain image information of the planting area; Step S2: Analyze the image information of the planting area to obtain the current vegetation features; Step S20: If current vegetation features exist, determine the current vegetation category based on the current vegetation features; Step S21: Determine the current vegetation growth status based on the current vegetation category and current vegetation characteristics, and record it in the preset vegetation information table; Step S22: If the current vegetation growth state is abnormal, obtain the coordinates of the abnormal vegetation growth area and output them; Step S23: When the current vegetation category does not exist, identify the abnormal vegetation area; Step S24: Obtain image information of abnormal vegetation areas based on abnormal vegetation areas, and analyze the image information of abnormal vegetation areas to determine the characteristics of vegetation branches and trunks; Step S25: Determine the soil adhesion to the branches based on the characteristics of vegetation branches; Step S26: When there is soil adhering to the branches, determine the abnormal growth and rooting characteristics; Step S27: When abnormal growth and rooting characteristics are present, output the vacancy region signal; Step S3: When there are no current vegetation features, output the vacancy area signal.

2. The plant status and space detection method based on image recognition according to claim 1, characterized in that, Also includes: Step S40: When there are no vegetation branch features or no abnormal growth and rooting features, obtain bud and leaf features based on the image information of the abnormal vegetation area. Step S41: When there are no buds or branches, output the empty area signal; Step S42: When there are bud and leaf characteristics and vegetation branch and trunk characteristics, determine the growth transition connection area based on the bud and leaf characteristics and vegetation branch and trunk characteristics; Step S43: If there is no growth transition connection region, output the vacancy region signal; Step S44: If a growth transition connection region exists, output an abnormal vegetation region signal.

3. The plant status and space detection method based on image recognition according to claim 2, characterized in that, It also includes a method for outputting a reasonable empty space region signal, which includes: Step S45: When there is a vacancy signal, search for vegetation characteristics in adjacent areas; Step S46: Determine the vegetation growth trend of adjacent areas based on the vegetation characteristics of adjacent areas; Step S47: If the vegetation growth trend in the adjacent area is an expansion trend, output a reasonable vacancy area signal; Step S48: If the vegetation growth trend in the adjacent area is not the same as the surrounding expansion growth trend, output the empty area signal; Step S49: When there is no vegetation growth trend in adjacent areas, determine the adjacent vegetation category based on the vegetation characteristics of adjacent areas; Step S50: Search the preset vegetation feature database based on the adjacent vegetation categories to determine the growth trend of adjacent vegetation, and then proceed to steps S47 to S48.

4. The plant status and space detection method based on image recognition according to claim 3, characterized in that, It also includes a method for outputting a reasonable vacancy area signal when there are no adjacent vegetation types, the method comprising: Step S450: Obtain the current example region; Step S451: Determine the empty space area and the number of empty space areas based on the current example area; Step S452: Determine the length of the empty space area based on the empty space area; Step S453: Determine the total length of vacant spaces based on the length of vacant areas, and determine the average length of vacant spaces based on the number of vacant areas. Step S454: Determine the deviation value of the occupied length based on the occupied length of the vacant area and the average occupied length of the vacant area; Step S455: If the deviation value of the occupied length falls within the preset reasonable deviation value range, output a reasonable empty space area signal.

5. The plant status and space detection method based on image recognition according to claim 4, characterized in that, It also includes a method for outputting a reasonable empty space area signal if the deviation value of the occupied length does not fall within the reasonable deviation value range. This method includes: Step S4550: Determine the current empty space region based on the current example region; Step S4551: Obtain the left and right adjacent vegetation regions corresponding to the current empty space region; Step S4552: Obtain the left adjacent region corresponding to the left adjacent vegetation region; Step S4553: Obtain the right adjacent region corresponding to the right adjacent vegetation region; Step S4554: If both the left and right adjacent areas are empty areas, output a reasonable empty area signal.

6. The plant status and space detection method based on image recognition according to claim 2, characterized in that, It also includes a method for performing a blowing operation, the method comprising: Step S420: When there are bud and leaf features but no vegetation branches or trunks, obtain the current environmental wind speed; Step S421: If the current ambient wind speed falls within the preset effective wind speed range, determine the initial bud position based on the characteristics of the bud and branches, and obtain the current bud position in real time to determine the change in bud position. Step S422: Determine the area occupied by buds based on the characteristics of buds and branches, and determine the range of changes in the location of abnormal buds based on the area occupied by buds; Step S423: If the change in bud position falls within the range of abnormal bud position changes, output the empty area signal; Step S424: If the change in bud position does not fall within the range of abnormal bud position changes, no empty area signal is output. Step S425: If the current ambient wind speed does not fall within the effective wind speed range, determine the blowing position and blowing force based on the initial bud point position, and form a blowing plan; Step S426: Control the preset vacancy detection device to perform the blowing operation according to the blowing scheme, and obtain the current bud position in real time to determine the amount of change in bud position, so as to execute steps S422 to S424.

7. The plant status and vacancy detection method based on image recognition according to claim 6, characterized in that, If the change in bud position falls within the range of abnormal bud position changes, the methods for outputting the vacancy region signal include: Step S4230: Obtain the image information of the initial position of the bud corresponding to the initial bud position; Step S4231: Determine the original bud features based on the initial position image information of the buds, and output them as bud and leaf features; Step S4232: If the original bud point feature exists, repeat steps S420 to S425.

8. The plant status and vacancy detection method based on image recognition according to claim 6, characterized in that, If there is no growth transition connection region, the methods for outputting the vacancy region signal include: Step S440: Determine the current number of branches and leaves and the current orientation of branches and leaves based on the characteristics of buds and branches; Step S441: If the current number of branches and leaves is not 1, determine the current branch orientation based on the characteristics of the vegetation branches; Step S4410: If the current number of branches and leaves is 1, output the empty area signal; Step S442: Determine the consistency quantity based on the current branch and leaf orientation and the current branch orientation; Step S443: If the number of consistent elements is equal to the current number of branches and leaves, do not output the empty area signal; Step S444: If the number of consistent vegetation is less than the current number of branches and leaves, determine the location of suspected abnormal vegetation; Step S445: Determine the blowing location and blowing force based on the suspected abnormal vegetation location to form a blowing plan, and execute the blowing operation according to the blowing plan; Step S446: Obtain the number of fallen branches and leaves during the blowing operation; Step S447: If the number of fallen branches and leaves is equal to the current number of branches and leaves, output the empty area signal.

9. The method for plant status and vacancy detection based on image recognition according to claim 8, characterized in that, The methods for performing the blowing operation include: Step S448: Obtain the location of the vacancy detection device and determine the current vacancy detection device number based on the air blowing position; Step S449: Determine the sliding amount, lifting and lowering amount, yaw angle adjustment amount, and pitch angle adjustment amount of the moving support based on the current empty space detection device number to form the empty space detection device movement parameters; Step S450: When there are no empty space detection device movement parameters, determine the current planting trough number based on the air blowing position; Step S451: Determine the valid empty space detection device number corresponding to the current planting trough number based on the current planting trough number, and re-determine the empty space detection device movement parameters; Step S452: Based on the movement parameters of the empty space detection device and the blowing force, a blowing plan is formed, and the blowing operation is performed according to the blowing plan.

10. A plant status and vacancy detection system based on image recognition, characterized in that, include: The acquisition module is used to acquire image information of the planting area; A memory for storing a program for a plant status and vacancy detection method based on image recognition as described in any one of claims 1 to 9; The processor loads and executes programs from memory.

Images