Early detection method and device for strawberry powdery mildew

Abstract

The invention discloses an early detection method and device for strawberry powdery mildew, and relates to the technical field of agricultural disease detection. The method comprises the steps of sample fixation and environment setting, ultraviolet excitation, image acquisition, image processing and analysis, and diagnosis result output. The device mainly comprises a supporting piece, an ultraviolet light source, an industrial camera and a computer. The ultraviolet light source emits exciting light and is matched with ultraviolet to penetrate through the filter to purify a spectrum; the industrial camera is provided with the combination of a polarizer and an ultraviolet filter, so that stray light and surface glare are effectively eliminated; the computer automatically analyzes fluorescence characteristics through an image processing algorithm. According to the present invention, the characteristic fluorescence response generated by the scab region under the ultraviolet excitation is utilized, and the fluorescence difference is quantified, such that the accurate diagnosis can be achieved during the disease incubation period, the early-stage, rapid and accurate diagnosis of the strawberry powdery mildew is achieved, and the innovative solution is provided for the crop disease treatment.

Description

Technical Field

[0001] This invention relates to the field of agricultural disease detection technology, specifically to a method and device for early detection of strawberry powdery mildew. Background Technology

[0002] Strawberries, an important economic crop in my country, are facing a serious threat from powdery mildew. This disease, caused by the pathogen *Podosphaera aphanis*, is highly contagious and spreads rapidly. In severe cases, the disease incidence rate can reach over 45% of leaves and over 50% of fruits, severely impacting strawberry yield, quality, and economic benefits. Powdery mildew is most likely to occur at temperatures between 15℃ and 25℃ and relative humidity between 40% and 80%. Outbreaks are more likely to occur in greenhouse cultivation environments due to high humidity and poor ventilation.

[0003] Currently, powdery mildew detection mainly relies on manual visual inspection and visible light imaging techniques. Manual inspection has significant limitations: early lesions are small and indistinct in color, easily confused with leaf pubescence or light spots, leading to a high rate of missed detection; in greenhouse or field environments, each plant and leaf needs to be inspected individually, which is time-consuming, labor-intensive, and inefficient; and the results are significantly affected by personal experience and subjective judgment, with different personnel having inconsistent standards for judging lesion characteristics, making it difficult to guarantee the accuracy and consistency of the detection.

[0004] Although technologies such as visible light, near-infrared imaging, and thermal imaging can identify lesions when they are obvious, these methods are not sensitive enough for early infection stages because the lesion characteristic signals are weak and easily interfered with by environmental factors, resulting in a high false negative rate.

[0005] Therefore, there is an urgent need to develop a new detection scheme that can overcome the problems of insufficient detection sensitivity, high false negative rate and low efficiency in existing technologies, and achieve early, rapid and accurate diagnosis of strawberry powdery mildew. Summary of the Invention

[0006] The purpose of this invention is to provide a method and device for early detection of strawberry powdery mildew, which solves the problems of insufficient detection sensitivity, high false negative rate and low efficiency in the prior art, and realizes early, rapid and accurate diagnosis of strawberry powdery mildew.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A method for early detection of strawberry powdery mildew includes the following steps: S1: Sample fixation and environment setup. Fix the sample with the front facing the camera unit at a predetermined distance, and set up a light-blocking environment to avoid interference from ambient light. S2: Ultraviolet excitation, using an ultraviolet light source to emit ultraviolet light in the range of 350-400nm, which is then filtered and purified before irradiating the leaves to excite the characteristic fluorescence response of the lesion area; S3: Image acquisition, acquiring fluorescence images through the camera unit and eliminating stray light and surface glare to ensure image quality; S4: Image processing and analysis, which transmits image data to the processing unit to perform preprocessing and feature extraction to quantify fluorescence differences; S5: Results output. Based on fluorescence differences, the location, size, and degree of infection of lesions are automatically identified, and diagnostic results are output.

[0008] Furthermore, the predetermined distance mentioned in S1 is 25-35cm to ensure that the sample is flattened in a stable posture within the imaging field of view of the camera unit.

[0009] Furthermore, the characteristic fluorescence response described in S2 refers to the fluorescence characteristics that early powdery mildew lesions exhibit, which are distinctly different from those of healthy tissue, when the sample is irradiated with ultraviolet light.

[0010] Furthermore, the elimination of stray light and surface glare described in S3 is achieved by combining a polarizing mirror and an ultraviolet filter. The polarizing mirror is used to eliminate polarized glare generated by reflection from the blade surface, and the ultraviolet filter is used to block stray light with wavelengths below 400nm.

[0011] Further, S4 includes: The obtained raw fluorescence images are preprocessed, including image denoising, brightness equalization and background correction, to improve image quality; Abnormal bright spot areas in fluorescence images were identified using image analysis algorithms; that is, early powdery mildew lesions appear as blue-green bright spots in fluorescence images. The shape and size of the lesions were determined through morphological processing and connected region analysis.

[0012] Another object of the present invention is to provide an early detection device for strawberry powdery mildew, which, when executed, implements the early detection method for strawberry powdery mildew according to any one of claims 1-5, comprising: Support components are used to support and secure the various components of the device; A white light source, fixed to the lower left side of the support, is used to provide visible light; An ultraviolet light source is fixed to the middle of the left side of the support and is used to emit ultraviolet light in the range of 350-400nm. An ultraviolet transmission filter is placed above the ultraviolet light source and fixed to the support to filter out light with a wavelength greater than 390nm. An industrial camera, fixed to the center of the support, has its lens aperture set to enable clear imaging of the sample within a range of 25-35cm. Polarizing filters and ultraviolet filters are installed in front of the lens of an industrial camera to filter stray light and eliminate glare reflected from the sample surface; The light source controller is electrically connected to the white light source, ultraviolet light source, and industrial camera to control the switching, luminous intensity, and synchronous triggering of the light source; An expandable connector is provided to connect to the device for connecting external expansion devices or additional sensors. The bracket is fixedly attached to the bottom of the support component and is used to fix the support component to the external mounting bracket; The computer, through an expandable interface, communicates with industrial cameras and light source controllers to control the image acquisition process and to process and identify lesions.

[0013] Furthermore, the ultraviolet light source comprises an array of multiple high-power ultraviolet LEDs, with a central emission wavelength of 365nm, and is arranged at an angle relative to the optical axis of the industrial camera.

[0014] Furthermore, the support is configured to fix the sample to be tested within a range of 25-35cm from the industrial camera.

[0015] Furthermore, the tilt angle is 15-45 degrees.

[0016] Furthermore, the application of an early detection method for strawberry powdery mildew in the detection of crops susceptible to powdery mildew.

[0017] In summary, the present invention has at least one of the following beneficial technical effects: By using ultraviolet-induced fluorescence differences as a detection signal, powdery mildew lesions can be accurately identified in the early stages that are difficult to observe with the naked eye, thus improving detection sensitivity and buying valuable time for disease prevention. The camera unit automatically acquires images and combines them with the processing unit's algorithm for recognition. The detection speed is fast and the efficiency is high, making it suitable for real-time monitoring of large-area crops. Systematic image acquisition and processing avoids the influence of human experience differences on the results, and the results of each test are standardized, which greatly reduces the false negative rate and the false positive rate, and improves the reliability of the test. The device has a reasonable structure, is easy to operate, and is highly expandable. It is easy to integrate into automated platforms such as intelligent greenhouse monitoring or field inspection to achieve remote real-time disease early warning and precise prevention and control, and has high value for promotion and application.

[0018] In summary, this invention solves the problems of insufficient detection sensitivity, high false negative rate, and low efficiency in the prior art, and realizes early, rapid and accurate diagnosis of strawberry powdery mildew, providing an innovative solution for crop disease management. Attached Figure Description

[0019] Figure 1 A schematic diagram of the fluorescence spectrum characteristics of strawberry powdery mildew; Figure 2 This is a flowchart of the method of the present invention; Figure 3 This is a schematic diagram of the detection results of the present invention; Figure 4 This is a schematic diagram of the overall structure of the present invention; In the picture: 1-White light source, 2-Ultraviolet light source, 3-Ultraviolet transmission filter, 4-Light source controller, 5-Industrial camera, 6-Polarizing filter and ultraviolet filter, 7-Expandable port, 8-Bracket, 9-Computer, 10-Support component. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0021] Experimental studies on the early detection of strawberry powdery mildew have revealed that when strawberry leaves are irradiated with ultraviolet light, early powdery mildew lesions exhibit fluorescence characteristics that are distinctly different from healthy tissue. This provides a new physical signal basis for the early identification of lesions: specifically, the intensity of chlorophyll red fluorescence (red fluorescence at a wavelength of approximately 680 nm) in the lesion area is weakened, while a significantly enhanced fluorescence signal is produced in the blue-green band (excitation wavelength approximately 350–400 nm, emission wavelength approximately 420–480 nm). This fluorescence difference indicates that early infected and healthy leaves can be distinguished by ultraviolet-excited fluorescence signals, thus providing a new approach and technical method for the early detection of strawberry powdery mildew.

[0022] This invention is based on the characteristic fluorescence response of strawberry leaves under ultraviolet excitation. A fluorescence imaging system is constructed to excite strawberry leaves under test with ultraviolet light and acquire their fluorescence images. The images are then processed and analyzed to identify early powdery mildew lesions. This device utilizes the difference in ultraviolet-induced fluorescence to significantly distinguish lesions from healthy tissue, enabling sensitive capture of early lesions that are difficult to detect with the naked eye.

[0023] like Figure 1As shown, the fluorescence intensity distribution of strawberry leaves under different excitation and emission wavelengths can be represented by the excitation-emission matrix (EEM). Fluorescence spectral measurements of healthy leaves and leaves infected with powdery mildew revealed significant differences in their fluorescence responses: healthy leaves exhibited a strong chlorophyll-red fluorescence peak (typical emission wavelength around 680 nm), while this red fluorescence was significantly weakened at early-stage infected lesions. Simultaneously, a marked enhancement of fluorescence was observed in the blue-green band of infected lesions, corresponding to an excitation wavelength of approximately 360 nm and an emission wavelength concentrated in the 420–480 nm range. Figure 1 The results visually reflect the aforementioned differences: the lesion area exhibits higher intensity in the blue-green fluorescent region. This experimental result verifies the feasibility of using ultraviolet-excited fluorescence to detect early lesions, providing a theoretical basis for this invention.

[0024] This embodiment provides a method for early detection of powdery mildew in strawberries, such as... Figure 2 As shown, it includes the following steps: S1: Sample fixation and environment setup. Fix the sample with the front facing the camera unit at a predetermined distance, and set up a light-blocking environment to avoid interference from ambient light. S2: Ultraviolet excitation, using an ultraviolet light source to emit ultraviolet light in the range of 350-400nm, which is then filtered and purified before irradiating the leaves to excite the characteristic fluorescence response of the lesion area; S3: Image acquisition, acquiring fluorescence images through the camera unit and eliminating stray light and surface glare to ensure image quality; S4: Image processing and analysis, which transmits image data to the processing unit to perform preprocessing and feature extraction to quantify fluorescence differences; S5: Results output. Based on fluorescence differences, the location, size, and degree of infection of lesions are automatically identified, and diagnostic results are output.

[0025] The following is a detailed description with reference to specific embodiments.

[0026] S1: Sample fixation and environment setup. Fix the sample with its front facing the camera unit at a predetermined distance, and set up a light-blocking environment to avoid interference from ambient light.

[0027] The sample (strawberry leaf) is fixed in the detection position, ensuring the leaf remains in a stable posture and flat within the imaging field of view of the camera unit. This prevents the leaf from shaking or bending during the detection process, which could affect the imaging effect and ensure the repeatability and accuracy of the fluorescence image acquisition process. The strawberry leaf is fixed within a range of 25-35cm from the camera unit, preferably 30cm, to ensure that the camera can acquire a clear image of the leaf surface. Simultaneously, the entire detection process is conducted in a light-protected environment (e.g., in a dark room or at night) to avoid interference from ambient light on the fluorescence imaging.

[0028] S2: Ultraviolet excitation, using an ultraviolet light source to emit ultraviolet light in the range of 350-400nm, which is then filtered and purified before irradiating the leaves to excite the characteristic fluorescence response of the lesion area.

[0029] The ultraviolet (UV) light source emits UV light in the 350-400 nm range. Positioned 30 cm directly in front of the sample, it provides UV excitation light to the strawberry leaves. This source emits UV light covering the 350–400 nm wavelength range to fully excite the fluorescence response of the leaf tissue. It targets the characteristic excitation peaks of early powdery mildew lesions, ensuring a clear fluorescence signal in the diseased areas, thus improving detection sensitivity and specificity. The UV light source wavelength can also be extended to 305–405 nm. A UV transmission filter is positioned above the UV light source; its core function is to purify the excitation light. It filters out components with wavelengths greater than 390 nm from the UV light source output, ensuring that the UV light irradiating the strawberry leaves is mainly concentrated in the ideal excitation band of 350-400 nm to excite the characteristic fluorescence response in the lesion area. The characteristic fluorescence response refers to the fluorescence characteristics that early powdery mildew lesions exhibit, clearly distinguishing them from healthy tissue, when the sample is irradiated with UV light.

[0030] S3: Image acquisition, which acquires fluorescence images through the camera unit and eliminates stray light and surface glare to ensure image quality.

[0031] The camera unit employs an industrial camera to capture fluorescence images of leaves under ultraviolet light excitation from the front. The industrial camera consists of a camera and a lens, with the lens aperture set appropriately to ensure clear imaging of the leaf within a 25-35cm range. The industrial camera is aimed at the leaf through the lens; when ultraviolet light illuminates the leaf, the camera opens its shutter to capture the fluorescence emitted from the leaf surface. This imaging result is sent to the processing unit via a data cable. Since healthy leaves primarily produce red fluorescence while lesions exhibit blue-green fluorescence, the camera can simultaneously acquire fluorescence information at different wavelengths, aiding in subsequent differentiation between lesions and healthy tissue. A polarizing filter and an ultraviolet filter are combined and mounted in front of the industrial camera lens, providing dual filtering functionality. The ultraviolet filter is responsible for blocking stray light with wavelengths below 400nm (mainly the remaining excitation ultraviolet light), ensuring that only the fluorescence emitted by the leaves (wavelength greater than 400nm) can enter the camera; while the polarizing filter is used to eliminate polarized glare generated by reflection from the leaf surface. By setting a polarizing filter in front of the camera and selecting an appropriate polarization angle, this type of reflected glare entering the camera can be minimized, further highlighting the weak fluorescence signal of the lesions and significantly improving the signal-to-noise ratio and contrast of the image.

[0032] S4: Image processing and analysis, which transmits image data to the processing unit to perform preprocessing and feature extraction to quantify fluorescence differences.

[0033] The processing unit, such as a computer, communicates with the camera unit to control the image acquisition process and to process and identify lesions. The processing unit performs algorithms such as correction, noise reduction, and feature extraction on the acquired fluorescence images, enabling it to automatically identify the location and extent of early powdery mildew lesions, thus achieving rapid diagnosis and early warning of the disease.

[0034] In this invention, image processing and analysis are implemented based on the deep learning target detection algorithm YOLOv13. This algorithm employs an end-to-end detection framework, enabling simultaneous location and category identification of lesions with a single image input. Compared to traditional image processing techniques based on color thresholding and morphological analysis, this method offers higher recognition accuracy and robustness, and is particularly suitable for identifying the complex morphology and weak fluorescence signal features of early-stage strawberry powdery mildew lesions.

[0035] First, the fluorescence images acquired through the fluorescence imaging system undergo format conversion and size normalization to ensure that the image size meets the input requirements of the model structure. Common input sizes include 416×416 pixels or 640×640 pixels. Before inputting the images into the model, they are normalized to adjust the pixel values ​​to between 0 and 1. Simultaneously, during the training phase, data augmentation strategies such as random flipping, scaling, and color perturbation are introduced to improve the model's generalization ability and adaptability to changes in lesion morphology under different lighting conditions.

[0036] The YOLOv13 model structure typically consists of a feature extraction network, a feature fusion network, and a detection output head. The feature extraction network is responsible for extracting multi-scale spatial feature information from fluorescence images; the feature fusion network uses a feature pyramid structure to fuse feature maps at different levels, enhancing the model's ability to recognize small-sized early lesions; and the detection output head outputs the position parameters, confidence scores, and class probabilities of candidate boxes at multiple scales.

[0037] After inputting a fluorescence image, the model outputs several candidate lesion regions. The prediction result for each candidate region includes the center coordinates, width, height, confidence score, and classification probability of the bounding box. All coordinates are normalized. During model training, three loss functions are minimized simultaneously to regress the overlap error of the bounding box position, the target confidence error, and the classification error. Commonly used regression loss functions include Intersection over Union (IoU) or Generalized Intersection over Union (GIoU), while cross-entropy loss is typically used for classification error.

[0038] The model was trained using a dataset of fluorescent images with manually labeled bounding boxes. Strawberry powdery mildew lesions in the images were labeled by experts based on fluorescence characteristics. After training, the model can automatically identify lesion regions in newly acquired fluorescent images. During the identification process, a non-maximum suppression algorithm is used to filter multiple predicted bounding boxes, eliminating overlapping areas and retaining only the lesion prediction box with the highest confidence.

[0039] The final identification results include parameters such as lesion location and confidence level. Figure 3 The method can visualize the results by adding pseudo-color annotations to the original image. The processing unit calculates the ratio between the identified lesion areas and the overall leaf area, outputting quantitative indicators such as infection ratio and lesion density, providing a reference for early intervention and precise control of strawberry powdery mildew. This method features high accuracy, strong anti-interference ability, and wide adaptability, enabling high-throughput, automated identification of early-stage strawberry powdery mildew lesions.

[0040] S5: Results output. Based on fluorescence differences, the location, size, and degree of infection of lesions are automatically identified, and diagnostic results are output.

[0041] The processing unit marks the identified lesion locations on the original image, generating a visualized detection result. Based on the detection result, the processing unit determines whether the plant is infected according to the YOLO recognition model's results and confidence level. Then, it determines the degree of infection based on the percentage of infected areas relative to the total number of leaves. This method considers leaves with less than 5% infected as early stage, 5%-20% as mid-stage, and more than 20% as late stage. The final result is then displayed to the user or uploaded to a cloud-based farm management system for remote monitoring and early warning.

[0042] This embodiment provides a device for early detection of strawberry powdery mildew, such as... Figure 4 As shown, it includes a white light source 1, an ultraviolet light source 2, an ultraviolet transmission filter 3, a light source controller 4, an industrial camera 5, a polarizing filter and an ultraviolet filter 6, an expandable port 7, a bracket 8, a computer 9, and a support component 10.

[0043] Support component 10 is used to support and fix the overall structure of the device. Support component 10 is a structural component of the device, responsible for ensuring the stability and reliability of the system during the detection process. Its design fully considers the needs of practical application scenarios, such as field or greenhouse environments. It is used to fix and support the overall structure of the device, including components such as the light source, camera, and filter module, preventing the imaging quality from being affected by shaking or external interference during the detection process.

[0044] White light source 1, fixed to the lower left side of support 10, is used to provide visible light. White light source 1 can acquire ordinary visible light images, facilitating the observation of visible light disease characteristics. This light source does not operate simultaneously with ultraviolet light source 2.

[0045] An ultraviolet light source 2, fixed to the middle left side of the support 10, emits ultraviolet light in the range of 350-400 nm. Positioned 30 cm directly opposite the sample, the ultraviolet light source 2 provides ultraviolet excitation light to the strawberry leaves. This light source emits ultraviolet light covering a wavelength range of 350–400 nm to fully excite the fluorescence response of the leaf tissue. It targets the characteristic excitation peaks of early powdery mildew lesions in strawberries, ensuring a clear fluorescence signal at the diseased site, thereby improving the sensitivity and specificity of detection. In this embodiment, the ultraviolet light source 2 is an array of multiple high-power ultraviolet LEDs with a central emission wavelength of 365 nm, arranged at an angle relative to the optical axis of the industrial camera 5. To improve the uniformity of excitation and reduce direct reflection, the ultraviolet light source 2 can be positioned at a certain angle relative to the optical axis of the industrial camera 5 to irradiate the leaves. This ensures sufficient ultraviolet irradiation of the entire leaf while preventing excessive specular reflection light from directly returning to the camera lens. The angle of inclination is 15-45 degrees, preferably 30 degrees. In addition, the UV light source band 2 can also be extended to 305–405nm.

[0046] The ultraviolet transmission filter 3, positioned above the ultraviolet light source 2 and fixed to the support 10, has the core function of purifying the excitation light. Specifically, it is designed to filter out components with wavelengths greater than 390nm in the output of the ultraviolet light source 2, ensuring that the ultraviolet light illuminating the strawberry leaves is mainly concentrated in the ideal excitation band of 350-400nm. This prevents non-target wavelength light from interfering with subsequent fluorescence imaging from the source.

[0047] Industrial camera 5 is used to acquire fluorescence images of leaves under ultraviolet light excitation from the front. Industrial camera 5 includes a camera and lens, with the lens aperture set appropriately to ensure clear imaging of the leaves within a 25-35cm range. A high-sensitivity color industrial camera is preferred to record the color and intensity distribution of the fluorescence signal. Industrial camera 5 is aimed at the leaf through its lens; when the ultraviolet light source 2 illuminates the leaf, the camera opens its shutter to capture the fluorescence emitted from the leaf surface. This imaging result is connected to computer 9 via a USB cable. Since healthy leaves primarily produce red fluorescence while lesions exhibit blue-green fluorescence, the color camera can simultaneously acquire fluorescence information at different wavelengths, aiding in subsequent differentiation between lesions and healthy tissue. It should be noted that white balance calibration and exposure parameter adjustments can be performed on the camera before use to ensure that the acquired fluorescence images are clear and detailed. The use of industrial camera 5 avoids interference from subjective human factors, providing a reliable raw data foundation for subsequent lesion identification. In addition, industrial camera 5 can also employ other imaging devices, such as CMOS or CCD cameras, or ordinary cameras with fluorescence filters.

[0048] The light source controller 4 is electrically connected to the white light source 1, the ultraviolet light source 2, and the industrial camera 5, and is used to control the switching of the light sources, their luminous intensity, and synchronous triggering. In this embodiment, the light source controller 4 is also connected to the industrial camera 5 to achieve synchronous coordination between the light source and the camera. When the camera 5 is ready to expose, the light source controller 4 triggers the ultraviolet light source 2 to emit ultraviolet light pulses and adjusts the brightness intensity of the ultraviolet light as needed. Through this synchronous triggering mechanism, the output of the ultraviolet light source 2 remains consistent each time the camera takes a picture, thereby ensuring the comparability of brightness between different batches of images, which is beneficial for subsequent image processing and analysis. In addition, the light source controller 4 can automatically adjust the power of the ultraviolet light source 2 based on the leaf reflection characteristics or fluorescence intensity feedback to obtain the best imaging effect.

[0049] A polarizing filter and an ultraviolet filter 6 are mounted in front of the lens of the industrial camera 5, forming a combined filtering module that performs dual filtering functions. The ultraviolet filter blocks stray light with wavelengths below 400nm (mainly residual excitation ultraviolet light), ensuring that only the fluorescence emitted by the leaves (wavelengths greater than 400nm) can enter the camera. The polarizing filter, on the other hand, eliminates polarized glare generated by reflections from the leaf surface. By placing a polarizing filter in front of the camera and selecting an appropriate polarization angle, this type of reflected glare entering the camera can be minimized, further highlighting the weak fluorescence signal of the lesions and significantly improving the signal-to-noise ratio and contrast of the image. Through the dual action of the polarizing filter and the ultraviolet filter 6, the influence of background stray light on the fluorescence image is greatly reduced, enhancing the contrast of the lesion fluorescence signal relative to the background.

[0050] Expandable connector 7, connected to the device, is used to connect external expansion devices or additional sensors, such as additional light sources, environmental sensors (temperature and humidity), communication modules (Wi-Fi or Bluetooth), or actuators. It uses standardized interfaces (such as USB-C, RJ45, or customized multi-pin interfaces) to ensure compatibility and ease of connection. Expandable connector 7 is an intelligent expansion module for the device, designed to enhance system flexibility and future compatibility, meeting the needs of agricultural IoT.

[0051] The bracket 8, fixed below the support member 10, is used to fix the sample to be tested, specifically the strawberry leaf, in its position. This keeps the leaf in a stable posture and flattened within the imaging field of view, preventing it from shaking or bending during the testing process and affecting the imaging effect. This ensures the repeatability and accuracy of the fluorescence image acquisition process. The bracket 8 is configured to fix the sample to be tested (strawberry leaf) within a distance of 25-35cm from the industrial camera 5, preferably 30cm, ensuring that the camera can acquire a clear image of the leaf surface.

[0052] Computer 9 serves as the system's control and data processing center. It communicates with the industrial camera 5 and the light source controller 4 via expandable port 7, controlling the image acquisition process and processing and identifying lesions. Computer 9 performs algorithms such as correction, noise reduction, and feature extraction on the acquired fluorescence images, automatically identifying the location and extent of early powdery mildew lesions, enabling rapid diagnosis and early warning of the disease. On one hand, computer 9 controls the shooting parameters and triggering sequence of the industrial camera 5 through corresponding interfaces and can send commands to the light source controller 4 to synchronize the on / off of ultraviolet light. On the other hand, computer 9 receives the fluorescence image data acquired by the industrial camera 5, storing and analyzing it. In terms of software implementation, computer 9 first performs necessary preprocessing on the acquired raw fluorescence images, such as image denoising, brightness equalization, and background correction, to improve image quality. After preprocessing, computer 9 uses image analysis algorithms to identify abnormal bright spot areas in the fluorescence images. Because early powdery mildew lesions appear as bright blue-green spots in fluorescence images, while healthy leaves primarily show a dark red background, the computer can extract these blue-green fluorescent regions using methods such as color component analysis and threshold segmentation. Furthermore, morphological processing and connected component analysis can be used to determine the shape and size of the lesions. In addition, the computer can mark the identified lesion locations on the original image, generating visualized detection results. Finally, based on the image analysis results, the computer determines whether strawberry leaves are infected with powdery mildew and the degree of infection, and can display the results to the user via a monitor interface or upload the data to a cloud-based farm management system for remote monitoring and early warning.

[0053] First, the support frame 10 and imaging device are fixed at a suitable position relative to the sample to be tested (strawberry leaf) using the bracket 8. The front of the strawberry leaf faces the industrial camera 5, and the distance from the industrial camera is 30cm, ensuring that the strawberry leaf is stably flattened in the imaging field of view of the industrial camera 5, ensuring that the industrial camera 5 can acquire a clear image of the leaf surface. Then, the ultraviolet excitation light source 2 provides excitation light of 350-400nm, which is irradiated onto the strawberry leaf through the ultraviolet transmission filter 3, amplifying the fluorescence characteristics of the lesions. The reflected light is then filtered by a polarizing filter and an ultraviolet filter 6 to eliminate various irrelevant light interferences before entering the lens of the industrial camera 5. The image information is then transmitted to the computer 9 through the expandable interface 7. The computer 9 intelligently analyzes and identifies the lesions, achieving rapid, sensitive, and reliable disease detection. Simultaneously, the entire detection process is best performed in a shaded environment (e.g., in a dark room or at night) to avoid interference from ambient light on fluorescence imaging. Through the above structural design and workflow, the strawberry powdery mildew early detection method of the present invention can acquire reliable fluorescence image signals and accurately identify the location of lesions at the early stage of disease occurrence.

[0054] Compared with existing technologies, this invention can significantly reduce the false negative and false positive rates, buying valuable time for the prevention and control of strawberry powdery mildew. Furthermore, due to its reasonable structure, simple operation, and strong scalability, the device is suitable for deployment in automatic inspection systems in smart greenhouses or in mobile field detection equipment, thus having broad application prospects in actual production. The detection targets can also be extended to other crops susceptible to powdery mildew, such as tomatoes and cucumbers.

[0055] Embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0056] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0057] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0058] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0059] Contents not described in detail in this specification are prior art known to those skilled in the art. It is hereby indicated that the above description is intended to help those skilled in the art understand this invention, but does not limit the scope of protection of this invention. Any equivalent substitutions, modifications, improvements, or simplifications of the above descriptions that do not depart from the essential content of this invention fall within the scope of protection of this invention.

Claims

1. A method for early detection of powdery mildew in strawberries, characterized in that, Includes the following steps: S1: Sample fixation and environment setup. Fix the sample with the front facing the camera unit at a predetermined distance, and set up a light-blocking environment to avoid interference from ambient light. S2: Ultraviolet excitation, using an ultraviolet light source to emit ultraviolet light in the range of 350-400nm, which is then filtered and purified before irradiating the leaves to excite the characteristic fluorescence response of the lesion area; S3: Image acquisition, acquiring fluorescence images through the camera unit and eliminating stray light and surface glare to ensure image quality; S4: Image processing and analysis, which transmits image data to the processing unit to perform preprocessing and feature extraction to quantify fluorescence differences; S5: Results output. Based on fluorescence differences, the location, size, and degree of infection of lesions are automatically identified, and diagnostic results are output.

2. The method for early detection of powdery mildew in strawberries according to claim 1, characterized in that: The predetermined distance mentioned in S1 is 25-35cm to ensure that the sample is flattened in a stable posture within the imaging field of view of the camera unit.

3. The method for early detection of powdery mildew in strawberries according to claim 1, characterized in that: The characteristic fluorescence response described in S2 refers to the fluorescence characteristics that early powdery mildew lesions exhibit, which are distinct from healthy tissue, when the sample is irradiated with ultraviolet light.

4. The method for early detection of powdery mildew in strawberries according to claim 1, characterized in that: The elimination of stray light and surface glare described in S3 is achieved by combining a polarizing mirror and an ultraviolet filter. The polarizing mirror is used to eliminate polarized glare generated by reflection from the blade surface, and the ultraviolet filter is used to block stray light with wavelengths below 400nm.

5. The method for early detection of powdery mildew in strawberries according to claim 1, characterized in that: S4 includes: The obtained raw fluorescence images are preprocessed, including image denoising, brightness equalization and background correction, to improve image quality; Abnormal bright spot areas in fluorescence images were identified using image analysis algorithms; that is, early powdery mildew lesions appear as blue-green bright spots in fluorescence images. The shape and size of the lesions were determined through morphological processing and connected region analysis.

6. A device for early detection of strawberry powdery mildew, characterized in that, When the device is executed, it implements the method for early detection of strawberry powdery mildew as described in any one of claims 1-5, comprising: Support member (10) is used to support and fix the various components of the device; A white light source (1) is fixed to the lower left side of the support (10) to provide visible light; An ultraviolet light source (2) is fixed to the middle of the left side of the support (10) and is used to emit ultraviolet light in the range of 350-400nm. An ultraviolet transmission filter (3) is set above the ultraviolet light source (2) and fixed to the support (10) to filter out light with a wavelength greater than 390nm; An industrial camera (5) is fixed to the center of a support (10), and its lens aperture is set to enable clear imaging of the sample within a range of 25-35cm. A polarizing mirror and an ultraviolet filter (6) are installed in front of the lens of an industrial camera (5) to filter stray light and eliminate glare reflected from the sample surface; The light source controller (4) is electrically connected to the white light source (1), the ultraviolet light source (2) and the industrial camera (5) and is used to control the switching of the light source, the light intensity and the synchronous triggering. An expandable connector (7) is connected to the device for connecting external expansion devices or additional sensors; The bracket (8) is fixed below the support (10) and is used to fix the support (10) and the external mounting bracket; The computer (9) is connected to the industrial camera (5) and the light source controller (4) via the expandable port (7) to control the image acquisition process and to process and identify lesions.

7. The strawberry powdery mildew early detection device according to claim 6, characterized in that: The ultraviolet light source (2) comprises an array of multiple high-power ultraviolet LEDs, with a central emission wavelength of 365nm, and is arranged at an angle relative to the optical axis of the industrial camera (5).

8. The strawberry powdery mildew early detection device according to claim 6, characterized in that: The bracket (8) is configured to fix the sample to be tested within a range of 25-35 cm from the industrial camera (5).

9. The strawberry powdery mildew early detection device according to claim 7, characterized in that: The tilt angle is 15-45 degrees.

10. The application of the method for early detection of powdery mildew in strawberries according to any one of claims 1-5 in the detection of crops susceptible to powdery mildew.

Images