Sweet potato seedling cultivation and growth condition detection system and method based on image recognition
An image recognition-based system automates sweet potato seedling growth detection, addressing inefficiencies in manual observation by providing real-time feedback on growth and pest issues.
Patent Information
- Application Number
- CN202411132970.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-16
- Publication Date
- 2024-12-06
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Current methods for detecting sweet potato seedling growth conditions rely heavily on labor-intensive on-site observation, which is time-consuming and inefficient.
A system and method utilizing image recognition technology, including an image acquisition, processing, analysis, and output module, to automatically detect seedling growth conditions and health status, with features for abnormal growth detection and insect infestation identification.
Enhances work efficiency by providing timely feedback on seedling growth abnormalities and insect infestations through automated image analysis, optimizing management practices.
Smart Images

Figure CN119090942A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of sweet potato seedling cultivation, and in particular to a sweet potato seedling cultivation growth condition detection system and method based on image recognition. Background Art
[0002] Sweet potato is an important root vegetable, widely grown in many parts of the world, especially in tropical and subtropical regions. Sweet potato is rich in nutrients, especially carbohydrates, dietary fiber, vitamins and minerals, which help lower blood sugar levels and promote cardiovascular health. Sweet potato is an indispensable part of the traditional diet. The cultivation of sweet potato seedlings is an important part of sweet potato planting. The correct cultivation method can improve the survival rate and growth quality of the seedlings.
[0003] The cultivation of sweet potato seedlings requires meticulous management and suitable environmental conditions. From seed selection to sowing, management, fertilization and transplanting, each link affects the final yield and quality. At present, the detection of the growth status of sweet potato seedlings is mostly achieved through on-site observation by staff. This is labor-intensive, time-consuming and difficult to improve work efficiency. Summary of the invention
[0004] 1. Technical issues to be solved
[0005] In view of the shortcomings of the prior art, the present invention provides a sweet potato seedling cultivation and growth detection system and method based on image recognition, which has the advantages of automatically collecting images of sweet potato seedlings and detecting the growth conditions of sweet potato seedlings based on the acquired relevant data, and can also detect the health status of sweet potato seedlings, thereby solving the above-mentioned problems.
[0006] (II) Technical solution
[0007] To achieve the above object, the present invention provides the following technical solution: a sweet potato seedling cultivation and growth detection system based on image recognition, comprising an image acquisition module, an image processing module, an image analysis module, an output module and an early warning module;
[0008] The image acquisition module acquires the image of the sweet potato seedlings once every week through a high-definition camera, and the image acquisition module sends the seedling image to the image processing module through a network;
[0009] The image processing module obtains the image width Kd, the image height Gd, the image pixel value Xs(x, y), the seedling pixel height Gdym and the pixel RGB value Xs(R, G, B), the image processing module is preset with a scale factor R, the image processing module calculates the actual length Gdsj of the seedling according to the seedling pixel height Gdym and the scale factor R, the image processing module calculates the seedling growth rate Szsl according to the actual length Gdsj of the seedling at different times, the image processing module is preset with a global threshold T, the image processing module calculates the binarized image value Ezxs( The image processing module calculates the number of pixels in the leaf area Xssl according to the image width Kd, the image height Gd and the binarized image value Ezxs(x,y), the image processing module calculates the seedling leaf area Ypmj according to the number of pixels in the leaf area Xssl, the image processing module calculates the seedling leaf area growth rate Mjsl according to the seedling leaf area Ypmj at different times, the image processing module calculates the image hue Tpsx, the image saturation Tpbh and the image brightness Tpmd according to the pixel RGB value Xs(R,G,B), and the image processing module sends the acquired data to the image analysis module through the network;
[0010] The image analysis module is preset with a seedling growth rate threshold Szsl yz And the threshold value of seedling leaf area growth rate Mjsl yz The image analysis module compares the seedling growth rate Szsl and the seedling leaf area growth rate Mjsl with the seedling growth rate threshold Szsl yz And the threshold value of seedling leaf area growth rate Mjsl yz Performing a comparison, the image analysis module sends the comparison result to the output module through the network;
[0011] The image analysis module is preset with a maximum value Tpsx of the image hue max 、Minimum value of image hue Tpsx min , Maximum image saturation value Tpbh max , minimum image saturation value Tpbh min 、Maximum brightness of the image Tpmd max And the minimum brightness of the picture Tpmd min The image analysis module compares the image hue Tpsx with the maximum image hue Tpsx max And the minimum hue value of the image Tpsx min By comparison, the image analysis module compares the image saturation Tpbh with the maximum image saturation Tpbh max And the minimum saturation value of the image Tpbh minFor comparison, the image analysis module compares the image brightness Tpmd with the maximum image brightness Tpmd max For comparison and minimum brightness of the image Tpmd min Performing a comparison, the image analysis module sends the comparison result to the output module through the network;
[0012] When any value of the seedling growth rate Szsl and the seedling leaf area growth rate Mjsl is lower than the corresponding threshold, it indicates that the seedling growth is abnormal, and the output module sends the abnormal signal to the early warning module through the network;
[0013] When the image hue minimum value Tpsx min ≤ Image hue Tpsx≤ Maximum image hue Tpsx max And the minimum saturation value of the image is Tpbh min ≤Picture saturation Tpbh≤Maximum saturation of picture Tpbh max And the minimum brightness of the picture is Tpmd min ≤ Image brightness Tpmd≤ Maximum image brightness Tpmd max , indicating that the seedlings are infested with insect pests, and the output module sends the abnormal signal to the early warning module through the network;
[0014] The early warning module feeds back the abnormal signal to the staff.
[0015] Preferably, the actual length of the seedling Gdsj algorithm is expressed as follows:
[0016] Gdsj=Gdym*R
[0017] In the formula, Gdsj represents the actual length of the seedling, Gdym represents the pixel height of the seedling, and R represents the scale factor, that is, the ratio of the actual length of the reference object to its pixel length in the image.
[0018] Preferably, the seedling growth rate Szsl algorithm expression is as follows:
[0019]
[0020] In the formula, Szsl represents the seedling growth rate, Gdsj i+1 represents the actual length of the seedling in week i+1, Gdsj i represents the actual length of the seedling in the i-th week, and Sjjg represents the time interval, which is 1 week here.
[0021] Preferably, the algorithm expression of the binarized image value Ezxs(x,y) is as follows:
[0022]
[0023] In the formula, Ezxs(x,y) represents the binarized image value, Xs(x,y) represents the original pixel value at position (x,y) in the image, T represents the global threshold, 1 represents foreground, and 0 represents background. It means that the area where the original pixel value Xs(x, y) is greater than or equal to the global threshold T is the foreground, and the area where the original pixel value Xs(x, y) is less than the global threshold T is the background.
[0024] Preferably, the algorithm expression of the number of pixels in the leaf area Xssl is as follows:
[0025]
[0026] In the formula, Xssl represents the number of pixels in the leaf area, Kd represents the image width, that is, the number of vertical pixels in the image, Gd represents the image height, that is, the number of horizontal pixels in the image, and Ezxs(x,y) represents the image value after binarization. Indicates the accumulation of all pixels in the image.
[0027] Preferably, the seedling leaf area Ypmj algorithm expression is as follows:
[0028] Ypmj=Xssl*Xsmj
[0029] In the formula, Ypmj represents the seedling leaf area, Xssl represents the number of pixels in the leaf area, and Xsmj represents the actual area represented by each pixel;
[0030] The seedling leaf area growth rate Mjsl algorithm expression is as follows:
[0031]
[0032] In the formula, Mjsl represents the growth rate of seedling leaf area, Ypmj i+1 represents the leaf area of seedlings in week i+1, Ypmj i represents the leaf area of seedlings in week i, and Sjjg represents the time interval, which is 1 week here.
[0033] Preferably, the image hue Tpsx algorithm expression is as follows:
[0034]
[0035] In the formula, Tpsx represents the image hue, R represents the red component, G represents the green component, B represents the blue component, max(R,G,B) represents the maximum value among the red component, green component and blue component, and min(R,G,B) represents the minimum value among the red component, green component and blue component;
[0036] The picture saturation Tpbh algorithm expression is as follows:
[0037]
[0038] In the formula, Tpbh represents the saturation of the image, max(R,G,B) represents the maximum value of the red component, green component, and blue component, and min(R,G,B) represents the minimum value of the red component, green component, and blue component;
[0039] The image brightness Tpmd algorithm expression is as follows:
[0040] Tpmd=max(R,G,B)
[0041] In the formula, Tpmd represents the brightness of the picture, and max(R,G,B) represents the maximum value of the red component, the green component, and the blue component.
[0042] Preferably, the image analysis module compares the seedling growth rate Szsl and the seedling leaf area growth rate Mjsl with the seedling growth rate threshold Szsl yz And the threshold value of seedling leaf area growth rate Mjsl yz For comparison, when any value of the seedling growth rate Szsl and the seedling leaf area growth rate Mjsl is lower than the corresponding threshold, it means that the seedling growth is abnormal, and the output module sends the abnormal signal to the early warning module through the network.
[0043] Preferably, the image analysis module compares the image hue Tpsx with the maximum image hue Tpsx max And the minimum hue value of the image Tpsx min By comparison, the image analysis module compares the image saturation Tpbh with the maximum image saturation Tpbh max And the minimum saturation value of the image Tpbh min For comparison, the image analysis module compares the image brightness Tpmd with the maximum image brightness Tpmd max For comparison and minimum brightness of the image Tpmd min For comparison, the minimum hue value of the image is Tpsx min ≤ Image hue Tpsx≤ Maximum image hue Tpsx max And the minimum saturation value of the image is Tpbh min ≤Picture saturation Tpbh≤Maximum saturation of picture Tpbh max And the minimum brightness of the picture is Tpmd min ≤ Image brightness Tpmd≤ Maximum image brightness Tpmd max , indicating that the seedlings are infested with insects, and the output module sends the abnormal signal to the early warning module through the network.
[0044] The method for detecting the growth status of sweet potato seedlings based on image recognition comprises the following steps:
[0045] S1, the image acquisition module acquires the image of sweet potato seedlings once every week through a high-definition camera, and the image acquisition module sends the seedling image to the image processing module through the network;
[0046] S2, the image processing module obtains the image width, image height, image pixel value, seedling pixel height and pixel RGB value, the image processing module is preset with a scale factor, the image processing module calculates the actual length of the seedling according to the seedling pixel height and the scale factor, the image processing module calculates the seedling growth rate according to the actual length of the seedling at different times, the image processing module is preset with a global threshold, the image processing module calculates the binarized image value according to the image pixel value and the global threshold, the image processing module calculates the number of pixels in the leaf area according to the image width, image height and the binarized image value, the image processing module calculates the seedling leaf area according to the number of pixels in the leaf area, the image processing module calculates the seedling leaf area growth rate according to the seedling leaf area at different times, the image processing module calculates the image hue, image saturation and image brightness according to the pixel RGB value, and the image processing module sends the acquired data to the image analysis module through the network;
[0047] S3, a seedling growth rate threshold and a seedling leaf area growth rate threshold are preset in the image analysis module, the image analysis module compares the seedling growth rate and the seedling leaf area growth rate with the seedling growth rate threshold and the seedling leaf area growth rate threshold respectively, and the image analysis module sends the comparison result to the output module through the network;
[0048] S4, the image analysis module is preset with the maximum image hue, the minimum image hue, the maximum image saturation, the minimum image saturation, the maximum image brightness and the minimum image brightness. The image analysis module compares the image hue with the maximum image hue and the minimum image hue, the image analysis module compares the image saturation with the maximum image saturation and the minimum image saturation, the image analysis module compares the image brightness with the maximum image brightness and the minimum image brightness, and the image analysis module sends the comparison result to the output module through the network;
[0049] S5. When any value of the seedling growth rate and the seedling leaf area growth rate is lower than the corresponding threshold, it indicates that the seedling growth is abnormal, and the output module sends the abnormal signal to the early warning module through the network;
[0050] S6. When the minimum value of the image hue is ≤ the image hue ≤ the maximum value of the image hue and the minimum value of the image saturation is ≤ the image saturation ≤ the maximum value of the image saturation and the minimum value of the image brightness is ≤ the image brightness ≤ the maximum value of the image brightness, it indicates that the seedlings are infested with insects, and the output module sends an abnormal signal to the early warning module through the network;
[0051] S7. The early warning module feeds back the abnormal signal to the staff.
[0052] Compared with the prior art, the present invention provides a sweet potato seedling cultivation growth condition detection system and method based on image recognition, which has the following beneficial effects:
[0053] The present invention compares the seedling growth rate and the seedling leaf area growth rate with the seedling growth rate threshold and the seedling leaf area growth rate threshold respectively. When any value of the seedling growth rate and the seedling leaf area growth rate is lower than the corresponding threshold, it indicates that the seedling growth is abnormal. The image hue is compared with the maximum image hue value and the minimum image hue value. The image analysis module compares the image saturation with the maximum image saturation value and the minimum image saturation value. The image analysis module compares the image brightness with the maximum image brightness value and the minimum image brightness value. If the minimum image hue value is ≤ image hue ≤ maximum image hue value, the minimum image saturation value is ≤ image saturation ≤ maximum image saturation value, and the minimum image brightness value is ≤ image brightness ≤ maximum image brightness value, it indicates that the seedlings are infested with insects, so that timely feedback is given to the staff, and corresponding measures are taken to improve work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS
[0054] Figure 1 It is a schematic diagram of the system flow of the present invention;
[0055] Figure 2 It is a schematic diagram of the steps of the method of the present invention. DETAILED DESCRIPTION
[0056] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0057] See also Figure 1 The sweet potato seedling cultivation and growth detection system based on image recognition includes an image acquisition module, an image processing module, an image analysis module, an output module and an early warning module. The image acquisition module obtains the sweet potato seedling image once a week through a high-definition camera, and the image acquisition module sends the seedling image to the image processing module through the network;
[0058] The image processing module obtains the image width Kd, image height Gd, image pixel value Xs(x,y), seedling pixel height Gdym and pixel RGB value Xs(R,G,B). The image processing module is preset with a proportional factor R. The growth status of sweet potato seedlings can be determined by detecting the height and leaf area of the seedlings to help judge the growth status of the seedlings. At the same time, the color of the seedling leaves can be detected to help judge whether the seedlings are infested with insects. Management suggestions or alarms are automatically generated based on the detection results to guide the staff to take corresponding measures.
[0059] The image processing module calculates the actual length of the seedling Gdsj according to the seedling pixel height Gdym and the proportional factor R. The algorithm expression of the actual length of the seedling Gdsj is as follows:
[0060] Gdsj=Gdym*R
[0061] In the formula, Gdsj represents the actual length of the seedling, Gdym represents the pixel height of the seedling, and R represents the proportional factor, that is, the ratio of the actual length of the reference object to its pixel length in the image. The image processing module calculates the seedling growth rate Szsl according to the actual length of the seedling Gdsj at different times. The algorithm expression of the seedling growth rate Szsl is as follows:
[0062]
[0063] In the formula, Szsl represents the seedling growth rate, Gdsj i+1 represents the actual length of the seedling in week i+1, Gdsj i represents the actual length of the seedling in week i, Sjjg represents the time interval, which is 1 week here;
[0064] By calculating the seedling growth rate Szsl, we can better understand the growth status of plants, optimize planting management, improve variety selection, predict yields, and support scientific research and environmental monitoring;
[0065] The image processing module has a preset global threshold T. The image processing module calculates the binarized image value Ezxs(x, y) based on the image pixel value Xs(x, y) and the global threshold T. The algorithm expression of the binarized image value Ezxs(x, y) is as follows:
[0066]
[0067] In the formula, Ezxs(x,y) represents the binarized image value, Xs(x,y) represents the original pixel value at position (x,y) in the image, T represents the global threshold, 1 represents foreground, and 0 represents background. The area where the original pixel value Xs(x, y) is greater than or equal to the global threshold T is the foreground, and the area where the original pixel value Xs(x, y) is less than the global threshold T is the background. The image processing module calculates the number of pixels in the leaf area Xssl according to the image width Kd, image height Gd and the binarized image value Ezxs(x, y). The algorithm expression of the number of pixels in the leaf area Xssl is as follows:
[0068]
[0069] In the formula, Xssl represents the number of pixels in the leaf area, Kd represents the image width, that is, the number of vertical pixels in the image, Gd represents the image height, that is, the number of horizontal pixels in the image, and Ezxs(x,y) represents the image value after binarization. It means that all pixels in the image are accumulated. The image processing module calculates the seedling leaf area Ypmj according to the number of pixels in the leaf area Xssl. The algorithm expression of the seedling leaf area Ypmj is as follows:
[0070] Ypmj=Xssl*Xsmj
[0071] In the formula, Ypmj represents the seedling leaf area, Xssl represents the number of pixels in the leaf area, and Xsmj represents the actual area represented by each pixel. The image processing module calculates the seedling leaf area growth rate Mjsl according to the seedling leaf area Ypmj at different times. The algorithm expression of the seedling leaf area growth rate Mjsl is as follows:
[0072]
[0073] In the formula, Mjsl represents the growth rate of seedling leaf area, Ypmj i+1 represents the leaf area of seedlings in week i+1, Ypmj i represents the leaf area of seedlings in week i, Sjjg represents the time interval, which is 1 week here;
[0074] The seedling leaf area growth rate Mjsl can reflect the health status of the seedlings, identify slow-growing or damaged seedlings at an early stage, and take timely measures. At the same time, by analyzing the growth of leaf area, water, fertilizer and light can be reasonably allocated to improve growth efficiency.
[0075] The image processing module calculates the image hue Tpsx, image saturation Tpbh and image brightness Tpmd according to the pixel RGB value Xs (R, G, B). The algorithm expression of the image hue Tpsx is as follows:
[0076]
[0077] In the formula, Tpsx represents the image hue, R represents the red component, G represents the green component, B represents the blue component, max(R,G,B) represents the maximum value among the red component, green component and blue component, and min(R,G,B) represents the minimum value among the red component, green component and blue component;
[0078] The expression of the picture saturation Tpbh algorithm is as follows:
[0079]
[0080] In the formula, Tpbh represents the saturation of the image, max(R,G,B) represents the maximum value of the red component, green component, and blue component, and min(R,G,B) represents the minimum value of the red component, green component, and blue component;
[0081] The image brightness Tpmd algorithm expression is as follows:
[0082] Tpmd=max(R,G,B)
[0083] In the formula, Tpmd represents the brightness of the image, and max(R,G,B) represents the maximum value of the red component, the green component, and the blue component;
[0084] Through color space conversion, feature extraction and threshold setting, color information can be effectively used to determine whether there is an insect-infested area;
[0085] The image processing module sends the acquired data to the image analysis module through the network;
[0086] The image analysis module has a preset seedling growth rate threshold Szsl yz And the threshold value of seedling leaf area growth rate Mjsl yz The image analysis module compares the seedling growth rate Szsl and the seedling leaf area growth rate Mjsl with the seedling growth rate threshold Szsl yz And the threshold value of seedling leaf area growth rate Mjsl yz After comparison is performed, the image analysis module sends the comparison result to the output module through the network;
[0087] The image analysis module has a preset maximum image hue value Tpsx max 、Minimum value of image hue Tpsx min , Maximum image saturation value Tpbh max , minimum image saturation value Tpbh min 、Maximum brightness of the image Tpmd max And the minimum brightness of the picture Tpmd min The image analysis module compares the image hue Tpsx with the maximum image hue Tpsxmax And the minimum hue value of the image Tpsx min By comparison, the image analysis module compares the image saturation Tpbh with the maximum image saturation Tpbh max And the minimum saturation value of the image Tpbh min For comparison, the image analysis module compares the image brightness Tpmd with the maximum image brightness Tpmd max For comparison and minimum brightness of the image Tpmd min After comparison is performed, the image analysis module sends the comparison result to the output module through the network;
[0088] When any value of the seedling growth rate Szsl and the seedling leaf area growth rate Mjsl is lower than the corresponding threshold, it means that the seedling growth is abnormal, and the output module sends the abnormal signal to the early warning module through the network;
[0089] When the image hue minimum value Tpsx min ≤ Image hue Tpsx≤ Maximum image hue Tpsx max And the minimum saturation value of the image is Tpbh min ≤Picture saturation Tpbh≤Maximum saturation of picture Tpbh max And the minimum brightness of the picture is Tpmd min ≤ Image brightness Tpmd≤ Maximum image brightness Tpmd max , indicating that the seedlings are infested with insects, and the output module sends the abnormal signal to the early warning module through the network;
[0090] The early warning module feeds back abnormal signals to the staff.
[0091] See also Figure 2 The method for detecting the growth of sweet potato seedlings based on image recognition comprises the following steps:
[0092] S1, the image acquisition module acquires the image of sweet potato seedlings once every week through a high-definition camera, and the image acquisition module sends the seedling image to the image processing module through the network;
[0093] S2, the image processing module obtains the image width, image height, image pixel value, seedling pixel height and pixel RGB value, the image processing module is preset with a scale factor, the image processing module calculates the actual length of the seedling according to the seedling pixel height and the scale factor, the image processing module calculates the seedling growth rate according to the actual length of the seedling at different times, the image processing module is preset with a global threshold, the image processing module calculates the binarized image value according to the image pixel value and the global threshold, the image processing module calculates the number of pixels in the leaf area according to the image width, image height and the binarized image value, the image processing module calculates the seedling leaf area according to the number of pixels in the leaf area, the image processing module calculates the seedling leaf area growth rate according to the seedling leaf area at different times, the image processing module calculates the image hue, image saturation and image brightness according to the pixel RGB value, and the image processing module sends the acquired data to the image analysis module through the network;
[0094] S3, a seedling growth rate threshold and a seedling leaf area growth rate threshold are preset in the image analysis module, the image analysis module compares the seedling growth rate and the seedling leaf area growth rate with the seedling growth rate threshold and the seedling leaf area growth rate threshold respectively, and the image analysis module sends the comparison result to the output module through the network;
[0095] S4, the image analysis module is preset with the maximum image hue, the minimum image hue, the maximum image saturation, the minimum image saturation, the maximum image brightness and the minimum image brightness. The image analysis module compares the image hue with the maximum image hue and the minimum image hue, the image analysis module compares the image saturation with the maximum image saturation and the minimum image saturation, the image analysis module compares the image brightness with the maximum image brightness and the minimum image brightness, and the image analysis module sends the comparison result to the output module through the network;
[0096] S5. When any value of the seedling growth rate and the seedling leaf area growth rate is lower than the corresponding threshold, it indicates that the seedling growth is abnormal, and the output module sends the abnormal signal to the early warning module through the network;
[0097] S6. When the minimum value of the image hue is ≤ the image hue ≤ the maximum value of the image hue and the minimum value of the image saturation is ≤ the image saturation ≤ the maximum value of the image saturation and the minimum value of the image brightness is ≤ the image brightness ≤ the maximum value of the image brightness, it indicates that the seedlings are infested with insects, and the output module sends an abnormal signal to the early warning module through the network;
[0098] S7. The early warning module feeds back the abnormal signal to the staff.
[0099] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
Claims
1. A sweet potato seedling cultivation and growth detection system based on image recognition, characterized in that: It includes image acquisition module, image processing module, image analysis module, output module and early warning module; The image acquisition module acquires the image of the sweet potato seedlings once every week through a high-definition camera, and the image acquisition module sends the seedling image to the image processing module through a network; The image processing module obtains the image width Kd, the image height Gd, the image pixel value Xs(x, y), the seedling pixel height Gdym and the pixel RGB value Xs(R, G, B), the image processing module is preset with a scale factor R, the image processing module calculates the actual length Gdsj of the seedling according to the seedling pixel height Gdym and the scale factor R, the image processing module calculates the seedling growth rate Szsl according to the actual length Gdsj of the seedling at different times, the image processing module is preset with a global threshold T, the image processing module calculates the binarized image value Ezxs( The image processing module calculates the number of pixels in the leaf area Xssl according to the image width Kd, the image height Gd and the binarized image value Ezxs(x,y), the image processing module calculates the seedling leaf area Ypmj according to the number of pixels in the leaf area Xssl, the image processing module calculates the seedling leaf area growth rate Mjsl according to the seedling leaf area Ypmj at different times, the image processing module calculates the image hue Tpsx, the image saturation Tpbh and the image brightness Tpmd according to the pixel RGB value Xs(R,G,B), and the image processing module sends the acquired data to the image analysis module through the network; The image analysis module is preset with a seedling growth rate threshold Szsl yz And the threshold value of seedling leaf area growth rate Mjsl yz The image analysis module compares the seedling growth rate Szsl and the seedling leaf area growth rate Mjsl with the seedling growth rate threshold Szsl yz And the threshold value of seedling leaf area growth rate Mjsl yz Performing a comparison, the image analysis module sends the comparison result to the output module through the network; The image analysis module is preset with a maximum value Tpsx of the image hue max 、Minimum value of image hue Tpsx min , Maximum image saturation value Tpbh max , minimum image saturation value Tpbh min 、Maximum brightness of the image Tpmd max And the minimum brightness of the picture Tpmd min The image analysis module compares the image hue Tpsx with the maximum image hue Tpsx max And the minimum hue value of the image Tpsx min By comparison, the image analysis module compares the image saturation Tpbh with the maximum image saturation Tpbh max And the minimum saturation value of the image Tpbh min For comparison, the image analysis module compares the image brightness Tpmd with the maximum image brightness Tpmd max For comparison and minimum brightness of the image Tpmd min Performing a comparison, the image analysis module sends the comparison result to the output module through the network; When any value of the seedling growth rate Szsl and the seedling leaf area growth rate Mjsl is lower than the corresponding threshold, it indicates that the seedling growth is abnormal, and the output module sends the abnormal signal to the early warning module through the network; When the image hue minimum value Tpsx min ≤ Image hue Tpsx≤ Maximum image hue Tpsx max And the minimum saturation value of the image is Tpbh min ≤Picture saturation Tpbh≤Maximum saturation of picture Tpbh max And the minimum brightness of the picture is Tpmd min ≤ Image brightness Tpmd≤ Maximum image brightness Tpmd max , indicating that the seedlings are infested with insect pests, and the output module sends the abnormal signal to the early warning module through the network; The early warning module feeds back the abnormal signal to the staff.
2. The sweet potato seedling cultivation and growth detection system based on image recognition according to claim 1, characterized in that: The actual length of the seedling Gdsj algorithm is expressed as follows: Gdsj=Gdym*R In the formula, Gdsj represents the actual length of the seedling, Gdym represents the pixel height of the seedling, and R represents the scale factor, that is, the ratio of the actual length of the reference object to its pixel length in the image.
3. The sweet potato seedling cultivation and growth detection system based on image recognition according to claim 2, characterized in that: The seedling growth rate Szsl algorithm expression is as follows: In the formula, Szsl represents the seedling growth rate, Gdsj i+1 represents the actual length of the seedling in week i+1, Gdsj i represents the actual length of the seedling in the i-th week, and Sjjg represents the time interval, which is 1 week here.
4. The sweet potato seedling cultivation and growth detection system based on image recognition according to claim 3, characterized in that: The algorithm expression of the binarized image value Ezxs(x,y) is as follows: In the formula, Ezxs(x,y) represents the binarized image value, Xs(x,y) represents the original pixel value at position (x,y) in the image, T represents the global threshold, 1 represents foreground, and 0 represents background. It means that the area where the original pixel value Xs(x, y) is greater than or equal to the global threshold T is the foreground, and the area where the original pixel value Xs(x, y) is less than the global threshold T is the background.
5. The sweet potato seedling cultivation and growth detection system based on image recognition according to claim 4, characterized in that: The algorithm expression of the number of pixels in the leaf area Xssl is as follows: In the formula, Xssl represents the number of pixels in the leaf area, Kd represents the image width, that is, the number of vertical pixels in the image, Gd represents the image height, that is, the number of horizontal pixels in the image, and Ezxs(x,y) represents the image value after binarization. Indicates the accumulation of all pixels in the image.
6. The sweet potato seedling cultivation and growth detection system based on image recognition according to claim 5, characterized in that: The seedling leaf area Ypmj algorithm expression is as follows: Ypmj=Xssl*Xsmj In the formula, Ypmj represents the seedling leaf area, Xssl represents the number of pixels in the leaf area, and Xsmj represents the actual area represented by each pixel; The seedling leaf area growth rate Mjsl algorithm expression is as follows: In the formula, Mjsl represents the growth rate of seedling leaf area, Ypmj i+1 represents the leaf area of seedlings in week i+1, Ypmj i represents the leaf area of seedlings in week i, and Sjjg represents the time interval, which is 1 week here.
7. The sweet potato seedling cultivation and growth detection system based on image recognition according to claim 6, characterized in that: The image hue Tpsx algorithm expression is as follows: In the formula, Tpsx represents the image hue, R represents the red component, G represents the green component, B represents the blue component, max(R,G,B) represents the maximum value among the red component, green component and blue component, and min(R,G,B) represents the minimum value among the red component, green component and blue component; The picture saturation Tpbh algorithm expression is as follows: In the formula, Tpbh represents the saturation of the image, max(R,G,B) represents the maximum value of the red component, green component, and blue component, and min(R,G,B) represents the minimum value of the red component, green component, and blue component; The image brightness Tpmd algorithm expression is as follows: Tpmd=max(R,G,B) In the formula, Tpmd represents the brightness of the picture, and max(R,G,B) represents the maximum value of the red component, the green component, and the blue component.
8. The sweet potato seedling cultivation and growth detection system based on image recognition according to claim 7, characterized in that: The image analysis module compares the seedling growth rate Szsl and the seedling leaf area growth rate Mjsl with the seedling growth rate threshold Szsl yz And the threshold value of seedling leaf area growth rate Mjsl yz For comparison, when any value of the seedling growth rate Szsl and the seedling leaf area growth rate Mjsl is lower than the corresponding threshold, it means that the seedling growth is abnormal, and the output module sends the abnormal signal to the early warning module through the network.
9. The sweet potato seedling cultivation and growth detection system based on image recognition according to claim 8, characterized in that: The image analysis module compares the image hue Tpsx with the maximum image hue Tpsx max And the minimum hue value of the image Tpsx min By comparison, the image analysis module compares the image saturation Tpbh with the maximum image saturation Tpbh max And the minimum saturation value of the image Tpbh min For comparison, the image analysis module compares the image brightness Tpmd with the maximum image brightness Tpmd max For comparison and minimum brightness of the image Tpmd min For comparison, the minimum hue value of the image is Tpsx min ≤ Image hue Tpsx≤ Maximum image hue Tpsx max And the minimum saturation value of the image is Tpbh min ≤Picture saturation Tpbh≤Maximum saturation of picture Tpbh max And the minimum brightness of the picture is Tpmd min ≤ Image brightness Tpmd≤ Maximum image brightness Tpmd max , indicating that the seedlings are infested with insects, and the output module sends the abnormal signal to the early warning module through the network.
10. A method for detecting the growth of sweet potato seedlings based on image recognition, characterized in that: The following steps are involved: S1, the image acquisition module acquires the image of sweet potato seedlings once every week through a high-definition camera, and the image acquisition module sends the seedling image to the image processing module through the network; S2, the image processing module obtains the image width, image height, image pixel value, seedling pixel height and pixel RGB value, the image processing module is preset with a scale factor, the image processing module calculates the actual length of the seedling according to the seedling pixel height and the scale factor, the image processing module calculates the seedling growth rate according to the actual length of the seedling at different times, the image processing module is preset with a global threshold, the image processing module calculates the binarized image value according to the image pixel value and the global threshold, the image processing module calculates the number of pixels in the leaf area according to the image width, image height and the binarized image value, the image processing module calculates the seedling leaf area according to the number of pixels in the leaf area, the image processing module calculates the seedling leaf area growth rate according to the seedling leaf area at different times, the image processing module calculates the image hue, image saturation and image brightness according to the pixel RGB value, and the image processing module sends the acquired data to the image analysis module through the network; S3, a seedling growth rate threshold and a seedling leaf area growth rate threshold are preset in the image analysis module, the image analysis module compares the seedling growth rate and the seedling leaf area growth rate with the seedling growth rate threshold and the seedling leaf area growth rate threshold respectively, and the image analysis module sends the comparison result to the output module through the network; S4, the image analysis module is preset with the maximum image hue, the minimum image hue, the maximum image saturation, the minimum image saturation, the maximum image brightness and the minimum image brightness. The image analysis module compares the image hue with the maximum image hue and the minimum image hue, the image analysis module compares the image saturation with the maximum image saturation and the minimum image saturation, the image analysis module compares the image brightness with the maximum image brightness and the minimum image brightness, and the image analysis module sends the comparison result to the output module through the network; S5. When any value of the seedling growth rate and the seedling leaf area growth rate is lower than the corresponding threshold, it indicates that the seedling growth is abnormal, and the output module sends the abnormal signal to the early warning module through the network; S6. When the minimum value of the image hue is ≤ the image hue ≤ the maximum value of the image hue and the minimum value of the image saturation is ≤ the image saturation ≤ the maximum value of the image saturation and the minimum value of the image brightness is ≤ the image brightness ≤ the maximum value of the image brightness, it indicates that the seedlings are infested with insects, and the output module sends an abnormal signal to the early warning module through the network; S7. The early warning module feeds back the abnormal signal to the staff.
Citation Information
Patent Citations
Crop growth situation monitoring Internet of Things system based on visual inspection
CN105574897A
Intelligent agriculture monitoring early warning system used for preventing and treating Or diseases and insect pests
CN108921849A
Rice growth vigor analysis system and method based on matching factors
CN114894791A
