Woody spring flower plant growth dynamic monitoring method based on low-altitude remote sensing technology
Through low-altitude drone remote sensing technology and deep learning algorithms, combined with image stretching and morphological analysis, the efficiency and accuracy of dynamic monitoring of growth of woody spring flower plants in traditional technology is solved, and efficient and accurate dynamic monitoring of growth is achieved.
Patent Information
- Application Number
- CN202410940199.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-15
- Publication Date
- 2024-11-15
AI Technical Summary
Traditional remote sensing technology cannot effectively monitor the growth dynamics of woody spring flower plants, relies on manual operations, consumes human resources and has poor timeliness.
Low-altitude drone remote sensing images are used to select the optimal image by stretching the image value ratio, and combined with SegNet deep learning image segmentation algorithm and morphological algorithm, high-precision dynamic monitoring of woody spring flower plants is achieved.
A large-scale, high-resolution and high-precision dynamic monitoring of woody spring flower plants is realized, which improves monitoring efficiency, and improves the classification accuracy of the deep learning algorithm through linear stretching of the ratio of image to original image, reducing data statistics time.
Smart Images

Figure CN118968345A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of woody spring flowering plant growth monitoring, and in particular relates to a woody spring flowering plant growth dynamic monitoring method based on low-altitude remote sensing technology. Background Art
[0002] Woody spring flowering plants refer to those woody flowers that bloom in spring. These plants are usually perennial, and their stems and roots are woody. Common woody spring flowering plants include cherry blossoms, magnolias, plum blossoms and other plants. Woody spring flowering plants have strong seasonal characteristics, and have their own unique flower colors and flowering periods in spring. They play an important role in landscaping and landscape arrangements, and are an important flower resource for spring viewing.
[0003] In recent years, deep learning has made great progress in image processing and data analysis. Deep learning has been applied to the identification and growth status assessment of crops such as rice and cotton. In view of the low resolution of traditional remote sensing, the inability to estimate the growth dynamics of woody spring flowering plants, poor timeliness, reliance on manual operation experience, and consumption of human resources, a deep learning algorithm for selecting the optimal image based on the ratio of image values before and after stretching was proposed based on low-altitude UAV remote sensing images. A method for extracting growth elements of woody spring flowering plants through morphological algorithms can realize large-scale and high-precision monitoring of the growth dynamic elements of woody spring flowering plants, providing data support for the management of woody spring flowering plants. Summary of the invention
[0004] The purpose of the embodiment of the present invention is to provide a method for dynamically monitoring the growth of woody spring flowering plants based on low-altitude remote sensing technology, aiming to solve the problems existing in the above-mentioned background technology.
[0005] The embodiment of the present invention is implemented as follows: a method for dynamically monitoring the growth of woody spring flowering plants based on low-altitude remote sensing technology, comprising:
[0006] Raw drone visible light image data;
[0007] Image stitching coordinate calibration;
[0008] UAV orthophoto visible light image data;
[0009] Linear stretch 5%;
[0010] Stretching results;
[0011] The interpolation of the original data pixel value and the stretched pixel value accounts for 5%-20% of the original data pixel value;
[0012] When the specific gravity is between 5% and 20%, the best image is obtained;
[0013] When the weight is not between 5% and 20%, it returns to the UAV orthophoto visible light image data and linearly stretches by 5%.
[0014] Preferably, the optimal image includes branch type sample selection, cherry blossom type sample selection and leaf type sample selection, based on the improved SegNet deep learning image segmentation algorithm (ratio of original data pixel value to stretched pixel value).
[0015] Preferably, it also includes ground object type classification, which is performed based on the results of an improved SegNet deep learning image segmentation algorithm based on (the ratio of original data pixel value to stretched pixel value).
[0016] Preferably, it also includes classification accuracy evaluation, and the classification accuracy evaluation is performed on the results after the ground feature type classification.
[0017] Preferably, a morphological target detection algorithm is also included, and the morphological target detection algorithm is calculated based on the results of the classification accuracy evaluation.
[0018] Preferably, it also includes statistics of classification results of individual plants, and the results obtained by the morphological target detection algorithm are used to perform statistics of classification results of individual plants.
[0019] Preferably, it also includes statistical results of all plants, and integrates the classification results of all individual plants to obtain statistical data of all plants.
[0020] Compared with the prior art, the method for dynamically monitoring the growth of woody spring flowering plants based on low-altitude remote sensing technology provided by the embodiment of the present invention has the following beneficial effects:
[0021] 1. For woody spring flowering plants, it can realize large-scale, high-resolution and high-precision growth dynamic monitoring, thereby improving the efficiency of monitoring;
[0022] 2. Introduce the ratio of linear stretched image to original image as a method to select the optimal image and improve the classification accuracy of deep learning algorithm;
[0023] 3. The number of image pixels of a single woody spring flowering plant after classification usually exceeds 10,000. The morphological algorithm is used to count the pixel statistics of each element of the single plant after classification, which greatly reduces the time required for data statistics. BRIEF DESCRIPTION OF THE DRAWINGS
[0024] Figure 1 A flow chart of a method for dynamically monitoring the growth of woody spring flowering plants based on low-altitude remote sensing technology provided in an embodiment of the present invention. DETAILED DESCRIPTION
[0025] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.
[0026] The specific implementation of the present invention is described in detail below in conjunction with specific embodiments.
[0027] like Figure 1 As shown, a structural diagram of a method for dynamically monitoring the growth of woody spring flowering plants based on low-altitude remote sensing technology provided by an embodiment of the present invention includes original UAV visible light image data; image stitching coordinate calibration; UAV orthophoto visible light image data; linear stretching of 5%; stretching result; the interpolation of the original data pixel value and the stretched pixel value accounts for a proportion of the original data pixel value between 5% and 20%; when the proportion is between 5% and 20%, the optimal image is obtained; when the proportion is not between 5% and 20%, it returns to between the UAV orthophoto visible light image data and the linear stretching of 5%.
[0028] The present application provides a method for dynamic monitoring of the growth of woody spring flowering plants based on low-altitude remote sensing technology. For woody spring flowering plants, large-scale, high-resolution and high-precision dynamic growth monitoring can be achieved, thereby improving the efficiency of monitoring; a ratio of a linearly stretched image to an original image is introduced as a method for selecting the optimal image, thereby improving the classification accuracy of a deep learning algorithm; the number of pixels in a single image of a classified woody spring flowering plant is usually more than 10,000, and a morphological algorithm is used to statistically calculate the pixel statistics of each element of a single plant after classification, thereby greatly reducing the time required for data statistics.
[0029] In one example of the present invention, the original visible light remote sensing image data obtained by the UAV is used as the basis. First, these data will go through the steps of image stitching and coordinate calibration to ensure the integrity and accuracy of the image in the same coordinate system; secondly, in the UAV orthophoto visible light image data processing stage, a key step is included - linear stretching of 5%, which stretches the image data to improve the contrast and visual effect of the image; finally, the process enters the comparison stage of the data pixel value and the stretched pixel value, and the interpolation processing of the pixel value is performed by comparing the original data pixel value with the stretched pixel value. This interpolation process needs to ensure that the proportion of the pixel value is between 5% and 20% to achieve the optimal selection of data and obtain the optimal image.
[0030] like Figure 1 As shown, as a preferred embodiment of the present invention, the optimal image includes branch type sample selection, cherry blossom type sample selection and leaf type sample selection, based on the improved SegNet deep learning image segmentation algorithm (ratio of original data pixel value to stretched pixel value);
[0031] It also includes the classification of land object types, which is performed based on the results of the improved SegNet deep learning image segmentation algorithm (ratio of original data pixel value to stretched pixel value);
[0032] It also includes classification accuracy evaluation, and the classification accuracy is evaluated after the results of land feature type classification.
[0033] In one example of the present invention, after obtaining the optimized image data, the classification and detection stage is entered. First, sample values of various objects are selected, and the different types of objects in the image are classified, such as tree branches, cherry blossoms, branches and leaves, etc., through the SegNet deep learning image segmentation algorithm. This step improves the classification accuracy of the image.
[0034] like Figure 1 As shown, as another preferred embodiment of the present invention, it also includes morphological target detection algorithm, single plant classification result statistics and all plant statistics results in sequence;
[0035] The results of the classification accuracy evaluation are used to calculate the morphological target detection algorithm. The results obtained by the morphological target detection algorithm are used to count the classification results of individual plants. The classification results of all individual plants are combined to obtain the statistical data of all plants.
[0036] In one example of the present invention, a morphological target detection algorithm is used to segment and detect individual plants. This step can accurately identify individual plants in the image, classify and count them, and finally count the results of all plants.
[0037] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above and that the present invention can be implemented in other specific forms without departing from the spirit or essential features of the present invention. Therefore, the embodiments should be considered exemplary and non-restrictive in all respects, and the scope of the present invention is defined by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims be included in the present invention.
[0038] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.
Claims
1. A method for dynamic monitoring of woody spring flowering plant growth based on low-altitude remote sensing technology. It is characterized in that include: Raw drone visible light image data; Image stitching coordinate calibration; UAV orthophoto visible light image data; Linear stretch 5%; Stretching results; The interpolation of the original data pixel value and the stretched pixel value accounts for 5%-20% of the original data pixel value; When the specific gravity is between 5% and 20%, the best image is obtained; When the weight is not between 5% and 20%, it returns to the UAV orthophoto visible light image data and linearly stretches by 5%.
2. A method for dynamically monitoring the growth of woody spring flowering plants based on low-altitude remote sensing technology according to claim 1, It is characterized in that The optimal image includes branch type sample selection, cherry blossom type sample selection and leaf type sample selection, based on the improved SegNet deep learning image segmentation algorithm (ratio of original data pixel value to stretched pixel value).
3. A method for dynamically monitoring the growth of woody spring flowering plants based on low-altitude remote sensing technology according to claim 2, It is characterized in that It also includes the classification of land object types, which is performed based on the results of the improved SegNet deep learning image segmentation algorithm (the ratio of the original data pixel value to the stretched pixel value).
4. A method for dynamically monitoring the growth of woody spring flowering plants based on low-altitude remote sensing technology according to claim 3, It is characterized in that It also includes classification accuracy evaluation, and the classification accuracy is evaluated after the results of land feature type classification.
5. A method for dynamically monitoring the growth of woody spring flowering plants based on low-altitude remote sensing technology according to claim 4, It is characterized in that It also includes a morphological target detection algorithm, and the results of the classification accuracy evaluation are used to calculate the morphological target detection algorithm.
6. A method for dynamically monitoring the growth of woody spring flowering plants based on low-altitude remote sensing technology according to claim 5, It is characterized in that It also includes statistics on the classification results of individual plants, and the results obtained by the morphological target detection algorithm are used to perform statistics on the classification results of individual plants.
7. A method for dynamically monitoring the growth of woody spring flowering plants based on low-altitude remote sensing technology according to claim 6, It is characterized in that It also includes the statistical results of all plants, and integrates the classification results of all individual plants to obtain the statistical data of all plants.
Citation Information
Patent Citations
A terrestrial plant ecological environment monitoring method based on multi-source remote sensing data fusion
CN109684929A
Mangrove forest ecological information intelligent extraction method based on unmanned aerial vehicle
CN112861837A
High-resolution remote sensing image forest grass change detection device and method based on deep learning
CN113657351A
Cited By
Field crop growth monitoring system and monitoring method based on low-altitude remote sensing
CN119774026A
