Remote sensing monitoring method and system for cultivated land protection

Abstract

The invention relates to the technical field of image processing, in particular to a remote sensing monitoring method and system for cultivated land protection, and the method comprises the following steps: obtaining a remote sensing image sequence of a cultivated land region at a differential time node, extracting pixel gray values of the same land parcel, arranging the pixel gray values according to time, calculating the ratio of a difference value between adjacent frames to an intra-frame gray average value, and obtaining a remote sensing image sequence; and marking positions continuously exceeding the fluctuation reference ratio, and generating a pixel region set. According to the method, a sampling frequency threshold optimization strategy is adopted, abnormal areas are screened in combination with albedo change rate, short-time fluctuation interference is reduced, time sequence consistency is enhanced, a solar incident angle and a terrain slope are introduced to calculate projection influence, a slope surface shielding area is eliminated, and space grid marking and boundary integrity judgment are adopted; the method improves the continuity of cultivated land distribution information, combines time sequence analysis, sampling optimization, physical characteristic compensation and spatial consistency correction, enhances the fine recognition of the dynamic change of cultivated land, and improves the monitoring precision and data reliability.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and particularly to a remote sensing monitoring method and system for cultivated land protection. Background Art

[0002] The technical field of image processing includes related technologies for collecting, analyzing, identifying, converting, storing and other processing operations on images using a computer or other electronic devices. The core contents of this technical field include image acquisition, image preprocessing, feature extraction, image classification and recognition, image compression and coding, etc. Image processing is widely used in many fields such as medical image analysis, industrial inspection, remote sensing monitoring, security monitoring, autonomous driving, etc. Its overall development depends on the continuous improvement of image sensor technology, image computing models, and image data processing methods. This field focuses on improving the clarity of images, enhancing the available information content of images, and effectively analyzing and interpreting the content of images through mathematical modeling and computing technologies.

[0003] Among them, the remote sensing monitoring method for cultivated land protection refers to a method for observing and analyzing the cultivated land scope and changes through remote sensing images. This patent theme addresses the remote sensing identification problem in cultivated land protection, covering technical matters such as obtaining cultivated land distribution information through multi-temporal remote sensing images, identifying cultivated land change areas using image difference analysis methods, and determining the nature of cultivated land by combining vegetation indices and land use types. This method generally uses means such as image contrast analysis, surface cover classification determination, and time series image change recognition to complete remote sensing monitoring operations, so as to accurately extract cultivated land information and monitor changes at the data level.

[0004] The existing technology relies on image comparison at fixed time nodes in cultivated land remote sensing monitoring, and fails to make full use of pixel-level gray scale sequence information, resulting in easy misjudgment caused by short-term environmental disturbances and affecting the stable identification of change areas. The sampling method is not adjusted according to the change frequency, resulting in data redundancy in some areas and information loss in some areas, affecting the balance of overall monitoring. Time series monitoring fails to effectively eliminate short-term abnormal fluctuations, resulting in a decrease in the credibility of local data and affecting the accurate judgment of the cultivated land change trend. Slope occlusion is not effectively identified, and surface information is missing in some areas due to shadow or reflection errors, affecting the precise demarcation of cultivated land boundaries. The spatial information expression method does not fully consider the integrity of plot boundaries, resulting in the fragmentation of information of some cultivated land plots and being unfavorable for the coherent analysis of large-scale cultivated land changes. These deficiencies affect data accuracy and consistency, and reduce the reliability of cultivated land monitoring results. Summary of the Invention

[0005] The purpose of the present invention is to solve the deficiencies existing in the prior art, and to propose a remote sensing monitoring method and system for cultivated land protection.

[0006] To achieve the above object, the present invention adopts the following technical solutions: A remote sensing monitoring method for cultivated land protection, comprising the following steps:

[0007] S1: Obtain a remote sensing image sequence of the cultivated land area at different time nodes, extract the pixel gray values of the same plot and arrange them in time, calculate the ratio of the difference between adjacent frames to the average gray value within the frame, mark the positions where the ratio continuously exceeds the fluctuation reference ratio, and generate a pixel area set;

[0008] S2: Call the pixel area set, set an interference frequency threshold to screen the frequent areas, and check the position and coverage relationship frame by frame, lower the priority of the pixels in the densely sampled area, and generate a reference value for the sampling priority of the time-series image;

[0009] S3: Call the reference value for the sampling priority of the time-series image, extract the multi-band albedo sequence of the corresponding area, set a time window and calculate the frame-by-frame change rate, compare the current frame change rate with the fluctuation value within the window, and obtain a set of sampling retention area segments;

[0010] S4: Call the set of sampling retention area segments, extract the solar incidence angle and slope parameters, calculate the illumination projection length and combine the change amplitude of the reflection intensity, screen the areas where the combined value exceeds the occlusion threshold, and obtain a set of positions of the occlusion exclusion image;

[0011] S5: Call the set of positions of the occlusion exclusion image, mark the remaining areas, track the extension state of the plot boundary, judge the integrity and coherence of the boundary, and generate distribution information of sampleable cultivated land image blocks.

[0012] As a further solution of the present invention, the pixel area set includes pixel positions, gray scale fluctuation ratios, and change continuity marks. The reference value for the sampling priority of the time-series image includes sampling areas, sampling frequency thresholds, and pixel sampling sorting. The set of sampling retention area segments includes albedo sequences, fluctuation rates, and jump thresholds. The set of positions of the occlusion exclusion image includes solar incidence angles, terrain slopes, projection lengths, and occlusion determination thresholds. The distribution information of sampleable cultivated land image blocks includes spatial indexes, position sets, and sampling coherent areas.

[0013] As a further solution of the present invention, the specific steps of S1 are:

[0014] S101: Obtain a remote sensing image sequence of the cultivated land area at different time nodes, extract and sort the pixel gray values of the same area, calculate the gray value difference between adjacent time nodes, and combine the average gray value within the image frame to calculate the ratio of the gray value difference of the pixel points to obtain a pixel gray ratio sequence;

[0015] S102: Based on the pixel grayscale ratio sequence, pixel positions whose grayscale ratio changes exceed a set reference ratio at multiple consecutive time nodes are screened, and grayscale fluctuation marked pixel point sets are obtained after aggregation;

[0016] S103: calling the grayscale fluctuation marked pixel point set, analyzing the spatial distribution characteristics of adjacent pixels, merging high fluctuation pixels, and forming a grayscale change region set.

[0017] As a further solution of the present invention, the grayscale difference calculation formula of adjacent time nodes is specifically:

[0018]

[0019] in, Represents the grayscale difference between the pixel (x, y) in the i-th and i+1-th frames. Represents the gray value of the pixel (x, y) in the i-th frame, μ i represents the average grayscale value of all pixels in the i-th frame image, N represents the total number of pixels in the image, Represents the algebraic sum of the grayscale value changes of all pixels in the image between the i-th to the i+1-th frame.

[0020] As a further solution of the present invention, the specific steps of S2 are:

[0021] S201: calling the grayscale change region set, setting a pixel frequency threshold, marking the region exceeding the threshold, checking the spatial position of the marked region by frame, analyzing the coverage, setting a threshold and screening the region with a higher frequency, and obtaining the distribution of high-frequency marked regions;

[0022] S202: Based on the distribution of the high-frequency annotated area, select a location in the densely sampled area to perform a sampling down-adjustment operation, while retaining the original sampling data level of the unannotated area to obtain an adjusted pixel sampling level;

[0023] S203: calling the adjusted pixel sampling level, recording the sampling order structure of the pixel points, arranging them in chronological order, and establishing a time-series image sampling priority result.

[0024] As a further solution of the present invention, the specific steps of S3 are:

[0025] S301: calling the time-series image sampling priority result, extracting the multi-band albedo sequence corresponding to the adjusted sampling area, setting the time window and calculating the albedo change rate between consecutive frames, sorting the change rate of the time node, and establishing the albedo change rate sequence;

[0026] S302: Based on the albedo change rate sequence, compare the difference between the change rate of the region in the current sampling frame and the fluctuation rate within the time window, and filter out the regions where the difference exceeds the albedo jump threshold to obtain the albedo stable regions;

[0027] S303: Invoke the albedo stable regions, extract the spatial distribution information of the retained regions, organize the regional segment data in time series, and generate a sampling retained region segment set.

[0028] As a further solution of the present invention, the specific formula for calculating the albedo change rate value is:

[0029]

[0030] where, represents the albedo change rate value of the pixel point (x, y) in the i-th frame image in band b, represents the albedo value of the pixel point (x, y) in band b in the i-th frame image, represents the average albedo value of all pixel points in band b in this frame of image, represents the albedo value of the pixel point (x, y) in band b in the selected reference time frame, and Δt ref,i represents the time interval between the current image frame and the reference image frame.

[0031] As a further solution of the present invention, the specific steps of S4 are:

[0032] S401: Invoke the sampling retained region segment set, extract the solar incidence angle and terrain slope values of the corresponding regions, calculate the projection length in the solar direction based on the incidence angle and slope values, identify the illumination projection range of the regions, and determine whether there is a slope occlusion condition to obtain the slope occlusion judgment result;

[0033] S402: Based on the slope occlusion judgment result, extract the change amplitude of the reflection intensity of the region, and perform a combined operation with the projection value to calculate the combined value of the region to obtain the occlusion influence combined value;

[0034] S403: Invoke the occlusion influence combined value, filter out the regions where the combined value exceeds the occlusion determination threshold, extract the corresponding image positions and organize the spatial information to generate a shaded exclusion image position set.

[0035] As a further solution of the present invention, the specific steps of S5 are:

[0036] S501: Invoke the set of image positions excluding occlusion, mark the distribution status of the remaining areas in the spatial grid, track the extended form of the corresponding plot boundaries in the image frame, analyze the spatial connection of the areas based on the continuity characteristics of the adjacent areas of the boundaries, and judge the integrity and coherence of the boundaries to obtain the boundary coherent areas;

[0037] S502: Based on the boundary coherent areas, screen the areas that meet the continuous determination conditions of the sampling structure, classify the qualified areas as candidate blocks, and obtain the candidate areas for sampling;

[0038] S503: Invoke the candidate areas for sampling, organize the spatial index information, and match the corresponding image positions to establish the distribution information of the sampled cultivated land image blocks.

[0039] A remote sensing monitoring system for cultivated land protection, comprising:

[0040] The temporal gray-scale change extraction module obtains the remote sensing image sequence of the cultivated land area at different time nodes, extracts and sorts the gray-scale values of the pixels of the same plot, calculates the gray-scale difference between adjacent time nodes, and makes a ratio judgment on the difference in combination with the gray-scale mean value within the image frame, marks the pixels that exceed the gray-scale fluctuation reference value in continuous multiple changes, and generates a pixel area set;

[0041] The high-frequency change area screening module invokes the pixel area set, sets the frequency threshold and marks the high-frequency areas, checks the spatial positions and coverage relationships of the marked areas, adjusts the sampling levels of the sampling dense areas, records the adjusted pixel sampling sorting, and generates a reference value for the sampling priority of the temporal image;

[0042] The albedo fluctuation screening module invokes the reference value for the sampling priority of the temporal image, extracts the multi-band albedo sequences of the adjusted sampling areas, calculates the albedo change rate within the time window, screens out the areas where the rate difference exceeds the albedo jump threshold, and obtains a set of sampling retention area segments;

[0043] The slope occlusion exclusion module invokes the set of sampling retention area segments, extracts the solar incident angle and terrain slope values of the corresponding areas, calculates the solar direction projection length, judges the occlusion conditions, and combines the regional reflection intensity change amplitude and the projection value to screen out the areas that exceed the occlusion determination threshold to obtain the set of image positions excluding occlusion;

[0044] The cultivated land image block extraction module invokes the set of image positions excluding occlusion, marks the spatial distribution of the remaining areas, tracks the plot boundaries in the image frame, judges the integrity and coherence of the boundaries, screens out the areas that meet the sampling structure continuity, outputs the spatial index and position set, and generates the distribution information of the sampled cultivated land image blocks.

[0045] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0046] In the present invention, by adopting a sampling frequency threshold optimization strategy, screening abnormal areas in combination with the albedo change rate, reducing short-term fluctuation interference, enhancing temporal consistency, introducing the calculation of projection influence of the solar incidence angle and terrain slope, eliminating the slope occlusion area, adopting spatial grid marking and boundary integrity determination, improving the coherence of cultivated land distribution information, and combining temporal analysis, sampling optimization, physical property compensation and spatial consistency correction, the fine recognition of the dynamic changes of cultivated land is enhanced, and the monitoring accuracy and data reliability are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0048] Figure 1 is a schematic diagram of the step flow of the present invention;

[0049] Figure 2 is a system module diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0050] The following will describe the technical solutions in the present invention with reference to the drawings.

[0051] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to represent examples, illustrations or explanations. Any embodiment or design solution described as "example" in the present invention should not be construed as being more preferred or more advantageous than other embodiments or design solutions. Exactly speaking, the use of the word "example" is intended to present concepts in a specific way. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one of the two.

[0052] In the embodiments of the present invention, "image" and "picture" can sometimes be used interchangeably. It should be noted that when their differences are not emphasized, the meanings they express are the same. "(of)", "corresponding" and "corresponding" can sometimes be used interchangeably. It should be noted that when their differences are not emphasized, the meanings they express are the same.

[0053] In the embodiments of the present invention, sometimes subscripts such as W1 may be written in a non-subscript form such as W1. When their differences are not emphasized, the meanings they express are the same.

[0054] To make the technical problems, technical solutions and advantages to be solved by the present invention clearer, the following will be described in detail with reference to the accompanying drawings and specific embodiments.

[0055] Please refer to Figure 1 , a remote sensing monitoring method for cultivated land protection, comprising the following steps:

[0056] S1: Obtain a sequence of remote sensing images of the cultivated land area at different time nodes, extract the gray values of the pixels of the same piece of land in the images and arrange them in chronological order, perform point-by-point operations on the gray differences between adjacent time nodes, judge the ratio of the differences in combination with the average gray value within the image frame, mark the pixel positions that exceed the gray fluctuation reference ratio in continuous changes, and generate a set of pixel regions;

[0057] S2: Call the set of pixel regions, set a frequency threshold and mark the regions with high frequencies, check the spatial positions and coverage relationships of the marked regions in the image frame by frame, perform sampling reduction operations on the positions in the densely sampled regions, while retaining the original sampling levels of the unmarked regions, and record the adjusted pixel sampling sorting structure to generate a reference value for the sampling priority of the time-series image;

[0058] S3: Call the reference value for the sampling priority of the time-series image, extract the multi-band albedo sequence corresponding to the adjusted sampling regions in the image frame, delimit a time window and calculate the albedo change rate between consecutive frames in the sequence, compare the difference between the change rate of each region in the current sampling frame and the fluctuation rate within the time window, and screen out the regions where the difference exceeds the albedo jump threshold to obtain a set of fragments of the sampling retention regions;

[0059] S4: Call the set of fragments of the sampling retention regions, extract the solar incidence angle and terrain slope values of the corresponding regions in the image, calculate the projected length in the solar direction for each region, judge whether there is a slope occlusion condition, and then screen out the regions where the combined value exceeds the occlusion determination threshold based on the combination of the change amplitude of the regional reflection intensity and the projected value to obtain a set of positions of the occlusion-excluded images;

[0060] S5: Call the set of positions of the occlusion-excluded images, mark the distribution states of the remaining regions in the spatial grid, track the extended shape of the corresponding plot boundaries in the image frame, judge the boundary integrity and coherence of the regions, select the regions that meet the continuous determination conditions of the sampling structure as candidate blocks, and output the spatial index and position set to generate the distribution information of the samplable cultivated land image blocks.

[0061] The set of pixel regions includes pixel positions, gray-scale fluctuation ratios, and change continuity markers. The time-series image sampling priority reference value includes sampling regions, sampling frequency thresholds, and pixel sampling sorting. The set of sampling retention region segments includes albedo sequences, fluctuation rates, and jump thresholds. The set of occlusion exclusion image positions includes solar incident angles, terrain slopes, projection lengths, and occlusion determination thresholds. The distributive information of the sampleable cultivated land image blocks includes spatial indices, position sets, and sampled coherent regions.

[0062] The specific steps of S1 are as follows:

[0063] S101: Obtain the remote sensing image sequence of the cultivated land area at different time nodes, extract and sort the pixel gray-scale values of the same region, calculate the gray-scale difference between adjacent time nodes, and combine the gray-scale mean within the image frame to calculate the ratio of the gray-scale differences of pixel points to obtain the pixel gray-scale ratio sequence;

[0064] The specific formula for the gray-scale difference between adjacent time nodes is as follows:

[0065]

[0066] Among them, represents the gray-scale difference between the pixel point (x, y) between the i-th and the (i + 1)-th frames of the image, represents the gray-scale value of the pixel point (x, y) in the i-th frame, μ i represents the average value of the gray-scale values of all pixel points in the i-th frame of the image, N represents the total number of pixels in the image, represents the algebraic sum of the gray-scale value changes of all pixel points in the image between the i-th and the (i + 1)-th frames;

[0067] This formula is used to calculate the gray-scale difference between pixel points in adjacent time nodes of the remote sensing image, integrating local gray-scale differences, the overall gray-scale change trend of the image, and the balanced correction term of all pixel changes. The parameter descriptions are as follows:

[0068] The image frames T? and T? correspond to the remote sensing images obtained on April 5, 2024, and May 5, 2024, respectively. The image size is 200 rows × 200 columns, with a total of 40,000 pixel points. The image type is a single-channel grayscale image, and the gray-scale value range is from 0 to 255. The gray-scale values of the pixel point (x = 100, y = 100) in the T? frame and the T? frame are obtained through image reading, and the gray-scale values are 122 and 139 respectively, that is:

[0069]

[0070] The average value of all pixel gray-scale values in the image frame T? is obtained by summing all pixels and then dividing by the total number of pixels. The total sum counted by the monitoring tool is 4,928,000, and the total number of pixels in the image is 40,000 pixel points:

[0071]

[0072] The total grayscale value of pixels in the image frame T? is 5260000, and we can calculate:

[0073]

[0074] The sum of all pixel grayscale changes from image frame T? to T? is the total difference monitoring system obtains 212000, substituting the number of image pixels N = 40000:

[0075]

[0076] Substituting the above into the formula:

[0077] The first item is the square difference of the grayscale change of the pixel itself:

[0078] (139-122) 2 =289;

[0079] The second term is the square difference of the image frame mean change:

[0080] (131.5-123.2) 2 =68.89;

[0081] Add the two items and take the square root to get the root value:

[0082]

[0083] The third item is the absolute value of the average difference of the pixel full image change:

[0084] |5.3|=5.3;

[0085] The final result is:

[0086]

[0087] The result shows that the grayscale difference of the pixel (100,100) between April and May 2024 compared with the overall image is 13.61, which means that the pixel has a moderate brightening trend in the local and global grayscale changes. This value is subsequently used to calculate the ratio with the image grayscale mean, and then construct the pixel grayscale ratio sequence. The larger the value, the more unstable the point is in the local frame-to-frame change, and the difference with the overall image is relatively significant.

[0088] S102: Based on the pixel grayscale ratio sequence, pixel positions whose grayscale ratio changes exceed a set reference ratio at multiple consecutive time nodes are screened, and grayscale fluctuation marked pixel point sets are obtained after aggregation;

[0089] Compare the ratio changes of each pixel at consecutive time nodes, and screen out the pixel points with obvious gray-scale changes at two or more consecutive time nodes. To define "obvious gray-scale changes", a ratio threshold needs to be set. For example, referring to the median and standard deviation of each ratio under a large-scale statistics, the threshold is set to 0.15. Traverse all pixel ratio sequences, and judge whether the ratio of each pixel in any two consecutive nodes is greater than 0.15. For example, the ratio sequence of pixel (100, 100) is 0.117, 0.167, 0.138. Since the second ratio is greater than 0.15 and the third ratio is slightly less than 0.15, although there is a node exceeding the threshold, it does not meet the condition of more than two consecutive nodes being greater than the threshold, so it is not selected. Then look at the ratio sequence of pixel (120, 120) which is 0.162, 0.185, 0.190. All three are greater than 0.15 and are consecutive time nodes, meeting the condition. Mark this pixel as a gray-scale fluctuation pixel and store it in the mark set, recording it in the form of a two-dimensional coordinate array. Assuming that the total number of pixels in the image is 40,000, finally about 1,200 pixels meet the conditions, forming a set of gray-scale fluctuation marked pixel points. In practical applications, such fluctuation pixel points often appear concentrated at the image nodes corresponding to the rapid growth or harvesting period of crops, corresponding to the crop replacement or management behaviors inside the farmland, and have practical reference for the entire image scene.

[0090] S103: Call the set of gray-scale fluctuation marked pixel points, analyze the spatial distribution characteristics of adjacent pixels, and merge the high-fluctuation pixel points to form a set of gray-scale change regions;

[0091] It is necessary to further identify its spatial distribution and determine whether a contiguous area is formed. The method is to traverse each pixel position in the marked set one by one, check the positions of its 8 adjacent pixels around it, and confirm whether they also exist in the gray-scale fluctuation set. For example, for the pixel (120, 120), check whether at least three pixels among the eight positions of (119, 119), (119, 120), (119, 121), (120, 119), (120, 121), (121, 119), (121, 120), and (121, 121) around it are also in the set. If this condition is met, the pixel and its neighbors are regarded as a preliminary aggregation unit, and multiple pixels that meet this condition are aggregated and grouped into the same group to form a larger area. By scanning the image in the horizontal and vertical directions, these adjacency relationships are continuously searched, and a unique number is assigned to each connected area to form a gray-scale change area set. For example, finally, 6 change areas are obtained, and the number of pixels contained in each area is 85, 74, 60, 52, 39, and 25 respectively. Through area statistics, it can be known that the proportions they account for in the image are 0.42%, 0.37%, 0.3%, 0.26%, 0.2%, and 0.13% respectively. Among them, the first two areas show a rectangular aggregation trend, which may reflect the crop replacement process inside large cultivated lands, and the latter several areas show a scattered dot-like aggregation, which may be related to local farmland management. The entire area set finally forms a gray-scale change area set, which serves as the basic data for subsequent analysis of the area change trend.

[0092] The specific steps of S2 are as follows:

[0093] S201: Call the gray-scale change area set, set the pixel frequency threshold, mark the areas exceeding the threshold, check the spatial positions of the marked areas frame by frame, analyze the coverage range, set the threshold and screen the areas with higher frequencies, and obtain the distribution of high-frequency marked areas;

[0094] During the calling stage, it is necessary to traverse and process the set of regions that have been marked with grayscale change annotations. For each region in the set, calculate the number of times it is marked as a changed region at all time nodes, that is, count the frequency of each region being recognized as a changed region in the image frame sequence. For example, if region Z1 appears 4 times in 5 frames of images, its frequency is 4. When setting the pixel frequency threshold, it is necessary to set it in combination with the overall distribution of the frequencies of all regions. First, the maximum value, minimum value, and average value of the frequencies of all regions can be counted. For example, the frequency values of all regions are between 1 and 6, and the average value is 3.2. The frequency threshold can be set as an integer greater than the average value, that is, set to 4, which means that only when a region is marked as a changed region at least 4 times in the image sequence, it is considered that the region has sufficiently stable change characteristics. Mark the regions with a frequency greater than or equal to 4. Check each frame of the image in the image sequence one by one. The pixel positions of the marked regions in the image are recorded as two-dimensional coordinates. For example, region Z1 contains pixel positions (120, 120), (121, 120), (120, 121), (121, 121). Check whether these positions are within the image boundary range and there is a marking record in each frame. If there are offset or discontinuous regions, they can be marked as position drift regions. Further, calculate the pixel coverage range of each high-frequency region in each frame, that is, calculate the number of all marked pixels in the region, and use this as an area statistical indicator. Compare the coverage ranges of all high-frequency regions numerically, and filter out the regions with an area greater than the set lower limit value. For example, the minimum pixel coverage is set to 50 pixels. If a region only covers 40 pixels, then this region is excluded. Finally, the regions that meet the conditions in terms of both frequency and area are retained as high-frequency marked regions, and then the spatial distribution of these regions is statistically sorted. For example, regions Z1, Z3, and Z5 are grouped into the same position band to form a group of high-frequency distribution regions, and finally the high-frequency marked region distribution is output.

[0095] S202: Based on the high-frequency marked region distribution, perform a sampling reduction operation at selected positions in the sampling dense region, and at the same time retain the original sampling data level of the unmarked regions to obtain the adjusted pixel sampling level;

[0096] First, it is necessary to confirm which regions are located in the sampling-dense areas of the image. The identification of the sampling-dense areas is completed by counting the number of sampling points per unit area in each frame of the image. For example, the image is divided into 10×10 grid blocks, and the number of labeled regions and sampling points in each grid block is counted. If the coverage ratio of high-frequency labeled regions in a certain grid is greater than 80%, and each pixel block contains more than 20 sampling points on average, then this grid block is determined as a sampling-dense area. Record the spatial positions and corresponding pixel indices of all such regions, and perform sampling reduction at these positions, that is, reduce the number of sampling points per unit area to a certain proportion of the original. For example, adjust from one sampling point per 64 pixels to one sampling point per 128 pixels. That is, if a certain region in an original frame of the image contains 20 sampling points, only 10 points are retained after adjustment, and the boundary pixels and region center pixels are preferentially retained according to the retention strategy, and the sampling points in the intermediate overlapping area are deleted. The sampling reduction operation is recorded in the pixel sampling level matrix. For non-high-frequency labeled regions, that is, unlabeled regions, no change in the sampling level is performed, and the original sampling structure is retained. For example, if a certain region has one sampling point per 128 pixels in the original image, this ratio is retained unchanged. Finally, integrate the sampling level data of different regions to obtain a complete adjusted pixel sampling level matrix containing high-frequency regions and unlabeled regions.

[0097] S203: Call the adjusted pixel sampling level, record the sampling sorting structure of the pixel points, organize them in chronological order, and establish the chronological image sampling priority result;

[0098] To organize the structure of this matrix, the sampling level values corresponding to each pixel point need to be sorted and organized in the chronological order of the image frames. For example, the sampling level values of pixel (100, 100) in image frames T1 to T6 are 1, 1, 2, 1, 1, 2 in sequence, indicating that it has been downsampled in the 3rd and 6th frames. Construct the sampling change trajectory of this pixel according to the chronological order. Repeat this operation to arrange the sampling levels of all pixel points in the order of the time axis to generate a sampling sorting structure table. Each row represents a pixel point, and each column represents the sampling level value of the corresponding frame. During the process of establishing the chronological image sampling priority, analyze the sampling level sequence of each pixel point. If it maintains a low-level sampling in most time frames, it is marked as a low-priority pixel point. If the sampling level remains high in multiple time frames, it is marked as a high-priority pixel point. For example, if pixel point (150, 150) maintains a high sampling level in 5 out of 6 frames, it is included in the high-priority sequence. Finally, form a sampling priority matrix arranged in time frames, record the sampling change trends of different pixels in the image sequence, and the constructed sampling priority structure is used as the basis for subsequent image data reading or processing.

[0099] The specific steps of S3 are as follows:

[0100] S301: Call the time-series image sampling priority result, extract the multi-band albedo sequence corresponding to the adjusted sampling area, set a time window and calculate the albedo change rate, organize the change rates for time nodes, and establish an albedo change rate sequence;

[0101] The specific formula for the albedo change rate value is as follows:

[0102]

[0103] Where, represents the albedo change rate value of the pixel point (x, y) in the i-th frame of the image in band b, represents the albedo value of the pixel point (x, y) in the i-th frame of the image in band b, represents the average albedo value of all pixel points in band b in this frame of the image, represents the albedo value of the pixel point (x, y) in band b in the selected reference time frame, Δt ref,i represents the time interval between the current image frame and the reference image frame;

[0104] This formula is used to quantify the local change rate of the albedo of a pixel point in a remote sensing image at a certain time frame i relative to the reference frame. Select the plot number as HJ-A12, and the central pixel coordinates of the corresponding area are (160, 160), and extract the albedo information of this pixel in the red band. According to the Sentinel-2 image data obtained on June 5, 2024, the albedo value after atmospheric correction is read as

[0105]

[0106] The image mean value in the red band of this image is obtained by summing the albedo values of all valid pixels and dividing by the number of valid pixel points. The total albedo value is 201835, and the number of pixels is 62500 pixels. Calculate:

[0107]

[0108] Set the reference time frame as the image obtained on May 20, 2024. This time point is an image after sowing and is used as the initial reference frame for judging the change of the surface albedo. Obtain the albedo value of this pixel point in the red band as:

[0109]

[0110] The time interval between the image shooting dates is obtained by subtracting the timestamps in the metadata. The interval between June 5, 2024 and May 20, 2024 is 16 days, and it is obtained that:

[0111] Δt ref,i = 16;

[0112] Substitute all the numerical values into the formula for calculation:

[0113] The first term is the deviation of the current frame from the image mean:

[0114] |0.317 - 0.322| = 0.005;

[0115] The second term is the albedo difference between the current frame and the reference frame:

[0116] |0.317 - 0.283| = 0.034;

[0117] The numerator part is the sum of the two:

[0118] 0.005 + 0.034 = 0.039;

[0119] The final change rate is:

[0120]

[0121] This result indicates that the albedo change rate of the pixel point (160, 160) in the red band is 0.0024375, with the unit of albedo value divided by days, which means that within a 16-day interval, the albedo of this pixel changes by approximately 0.0024 per day on average. This value is used to screen for stable or abnormal ground object areas when comparing with other pixels or regions in the follow-up. By sorting the sequences with other bands and pixels, a complete albedo change rate sequence can be formed for time series remote sensing analysis.

[0122] S302: Based on the albedo change rate sequence, compare the difference between the change rate of the region in the current sampling frame and the fluctuation rate within the time window, and screen out the regions where the difference exceeds the albedo jump threshold to obtain the albedo stable regions;

[0123] It is necessary to calculate the difference between the albedo change rate of each sampled area in the current image frame and the reference rate within the time window. When operating, first select the image frame to be analyzed, such as frame T4, extract the pixel set of a certain area Z1 in it, and then obtain the change rate of the corresponding pixels in this area during frames T3 to T5. For example, the rate of pixel (150, 150) in the red band is 0.0003 / day, and its average rate within the time window T2 to T4 is 0.001 / day, then the difference is 0.0007 / day. Make an absolute value judgment on this difference. If the difference is greater than the set albedo jump threshold, then eliminate this pixel point. The setting of the jump threshold refers to the standard deviation statistics of the rate differences in the entire image. For example, the standard deviation of the rate differences of all pixels in a certain band is 0.0005, and the jump threshold can be set to 0.0006, indicating that when the rate difference of a certain pixel point is greater than 0.0006, it is determined that the change is abnormal. If the number of eliminated pixels in a certain area exceeds 30% of the total number of pixels in this area, then the entire area is not retained. For example, area Z2 contains 100 pixel points, and the rate differences of 40 pixel points among them exceed the threshold, then area Z2 is excluded as a whole. On the contrary, if only 12 pixels in area Z1 have rate differences exceeding the threshold and the remaining 88 are stable, then area Z1 is determined as an albedo stable area. Finally, traverse all sampled frames and their corresponding areas, and output the set of areas that meet the above conditions, which is the albedo stable area.

[0124] S303: Call the albedo stable area, extract the spatial distribution information of the retained area, organize the area segment data in time series, and generate a set of sampled retained area segments;

[0125] Read the two-dimensional coordinate information of each pixel point in the area one by one, and aggregate the pixel points corresponding to the area determined to be stable in each frame into an image segment. For example, in frame T3, area Z1 contains pixels (100, 101), (101, 101), (102, 101), and in frame T4, it is (100, 101), (101, 101), (101, 102). Then, group and organize them into two area segments according to the time frame, and at the same time retain the time frame index and area number. On this basis, sort the segments in time series, that is, record the corresponding positions of each area number in the image frame in the order of T1, T2, T3... Tn. If a certain area number appears stably in consecutive frames, it is recorded as a continuously stable segment. If it is interrupted, it is treated as a separate segment. For example, area Z1 exists continuously in T2, T3, and T4, is missing in T5, and appears again in T6. Then, it is split into two area segments Z□a and Z□b corresponding to frames T2 to T4 and T6 respectively. When constructing the final segment set, organize all area segments into structured data items according to information such as frame index, spatial position, and number of pixels. For example, Z□a corresponds to 3 frames, contains 12 pixels, and the center position is (100, 100). Z□b corresponds to 1 frame and contains 9 pixels. Summarize all area segment information in the above format to generate a sampling retention area segment set.

[0126] The specific steps of S4 are as follows:

[0127] S401: Call the sampling retention area segment set, extract the solar incidence angle and terrain slope values of the corresponding area, calculate the projection length in the solar direction based on the incidence angle and slope values, identify the illumination projection range of the area, and judge whether there is a slope occlusion condition to obtain the slope occlusion judgment result;

[0128] Extract the spatial coordinate range and corresponding time frame index of each segment area one by one. For the pixel positions in each area, obtain the solar incidence angle and terrain slope value on the corresponding remote sensing image frame. The solar incidence angle data can be obtained by calculating based on the image capture time and solar azimuth angle. For example, the solar incidence angle corresponding to the image obtained at 10:00 am on June 21, 2023 is 35 degrees. The terrain slope value is calculated from the terrain elevation model data. For example, the slope value of a certain area is 20 degrees. Then, based on the relative angle between the incidence angle and the slope value, calculate the surface projection length in the solar direction, specifically converted into the shadow length of each pixel point through trigonometric relations. Under the conditions of a 30-degree incidence angle and a 20-degree slope, this angle is 10 degrees. If the height difference of this area is 2 meters, the projection length is approximately 11.4 meters, that is, it may correspond to 3 to 4 pixel lengths on the image. Match the projection length with the pixel coverage of the original area to determine whether there are some pixels blocked by the projection. If a pixel is within the projection area and its slope direction relative to the solar direction deviates from the incidence angle direction, it is judged that slope surface occlusion has occurred. Set the occlusion condition as: when the angle between the slope direction and the incident direction exceeds 90 degrees and falls within the projection coverage range, it is recorded as an occluded pixel. If pixel points (120, 120), (121, 120), (122, 120) in a certain area segment all meet this condition, it is considered that there is slope surface occlusion in this area. Summarize the judgment results of all area segments and output the slope surface occlusion judgment results corresponding to each area.

[0129] S402: Based on the slope surface occlusion judgment result, extract the variation range of the reflection intensity of the area, and perform a combined operation with the projection value to calculate the combined value of the area and obtain the combined value affected by occlusion;

[0130] Extract the reflection intensity values of each pixel in each area judged to be occluded in the multi-temporal image frames, and separately count the variation range of the reflection intensity of these pixel points. The calculation method of the variation range is the maximum reflection intensity value minus the minimum value. For example, the reflection intensities of pixel (120, 120) in frames T1 to T6 are 0.26, 0.28, 0.25, 0.29, 0.27, 0.26, then the variation range is 0.29 minus 0.25 equal to 0.04. Perform this statistical operation on all occluded pixels within each area to calculate the overall average variation range of the reflection intensity of the area. For example, there are 25 pixels in area Z1, and the average variation range is 0.045. Then, perform a combined operation with the projection length calculated in the previous step. The combined operation rule is the product of the two, which is used to measure the amplification of the reflection change intensity caused by occlusion. For example, the average projection length of area Z1 is 4 pixels, and the variation range is 0.045, then the combined value is 0.18. After completion, perform the same operation process on all areas to obtain their respective combined values affected by occlusion, and then bind the area number to the corresponding combined value to form a structured record for subsequent judgment.

[0131] S403: Call the occlusion influence combined value, filter out the areas where the combined value exceeds the occlusion determination threshold, extract the corresponding image positions, organize the spatial information, and generate a set of image positions for occlusion exclusion;

[0132] Perform screening processing on the combined values of each area in sequence. Set an occlusion determination threshold to divide the areas with significant and non-significant occlusion influence. This threshold is set according to the distribution of the combined values in the entire image. For example, if the maximum value of the combined values of all areas is 0.24, the average value is 0.12, and the standard deviation is 0.05, then the threshold can be set to 0.18 as the screening boundary. Compare the combined value of each area with 0.18 one by one. If the combined value is higher than this value, then this area is marked as an area with significant occlusion influence. Extract the spatial position information of this area in the image, including the starting coordinates, pixel range, and the index of the time frame it belongs to, and convert it into a coordinate structure recognizable by the image processing module. For example, record the occlusion area of area Z2 in the range of (110, 110) to (114, 114) in image frame T4 as T4_110_110_114_114. Organize all the image positions that meet the occlusion conditions in sequence to generate a set of image positions for occlusion exclusion. Each item in this set contains information such as the image number, spatial range, and area number, which can be used by the image processing system for subsequent area occlusion exclusion operations.

[0133] The specific steps of S5 are as follows:

[0134] S501: Call the set of image positions for occlusion exclusion, mark the distribution state of the remaining areas in the spatial grid, track the extended form of the corresponding plot boundaries in the image frame, analyze the spatial connection of the areas based on the continuity characteristics of the adjacent areas of the boundaries, and judge the integrity and coherence of the boundaries to obtain the boundary coherent areas;

[0135] Perform spatial grid division processing on the remaining unexcluded regions in the image frame. Use an equally spaced grid structure to divide the image into multiple sub-blocks of a fixed size. For example, each sub-block is 40×40 pixels. Mark whether each sub-block contains remaining regions. If there are pixel coordinates not in the occlusion exclusion set, the corresponding sub-block is marked as "valid"; otherwise, it is marked as "invalid". After completing the recording of the spatial distribution state, perform plot boundary recognition on the valid sub-blocks in each frame of the image, and extract the extended form of the plot boundary. When extracting, traverse the boundary connection relationship between sub-blocks in an 8-neighborhood manner to construct a plot contour graph, and identify whether there are abnormal situations such as breaks, interruptions, and abnormal extensions in the contour. For example, if three adjacent sub-blocks are located at positions (2,3), (2,4), and (3,4), and the middle sub-block is in an invalid state, it is regarded as a boundary break. Then, make a judgment based on the spatial continuity characteristics of the adjacent boundary regions. The continuity judgment basis is the valid state of adjacent sub-blocks and the coherence degree of the pixel structure. Set the continuity judgment criterion as follows: If the number of valid states in three adjacent sub-blocks ≥ 2, and the number of row or column pixels with overlapping boundary pixels exceeds 50%, it is judged as spatially continuous. If this condition is continuously satisfied for more than four sub-blocks in a certain region along the horizontal or vertical direction, it is considered that the boundary of this region has coherence. Analyze the boundary structure composed of all remaining regions frame by frame in this way, and finally form a set of regions that meet the boundary integrity and coherence conditions, and output them as boundary coherent regions.

[0136] S502: Based on the boundary coherent regions, screen the regions that meet the sampling structure continuous determination conditions, classify the regions that meet the conditions as candidate blocks, and obtain the candidate regions for sampling;

[0137] Check the continuity of the internal pixel sampling structure region by region. The continuity determination of the sampling structure needs to be analyzed based on the minimum distance between sampling points, the arrangement direction, and the coverage density. Set the continuous determination condition as follows: The distribution distance of sampling points along the main direction in the region is less than or equal to 80 pixels, and the number of pixels covered by all sampling points in the region is greater than 30% of the total number of pixels in this region. For example, in a region with a size of 3200 pixels (40×80 pixels), if the sampling points cover a total of 1100 pixels, the coverage rate is 34.4%, which meets the density condition. If these points are evenly distributed along the horizontal direction and the adjacent spacing is between 70 and 80 pixels, it meets the arrangement spacing condition, and this region is classified as a sampling structure continuous region. Extract all the boundary coherent regions that meet the above conditions, summarize them into a candidate block set, number these candidate blocks, and attach their spatial coordinate index, frame index, and pixel coverage range information. For example, candidate block C01 is located in frame T3 of the image, with the starting coordinates (120,80), a range of 40×80 pixels, a sampling point coverage ratio of 35%, and a sampling spacing of 75 pixels. Form a structured table record for all candidate blocks, and finally output the candidate regions for sampling.

[0138] S503: Call the sampleable candidate regions, organize the spatial index information, match the corresponding image positions, and establish the distribution information of the sampleable cultivated land image blocks;

[0139] Read the frame index, coordinate starting point, pixel range, and number information of each region, and organize this information into a unified spatial index structure. Each record contains a four-tuple of the image frame number, candidate block number, and spatial position (starting row, starting column, row range, column range). For example, the spatial index of candidate region C02 in frame T4 is T4_C02_100_120_40_40. Then, match these spatial indices with the position mapping of each frame in the original image data, map the index information to the specific pixel block positions in the image matrix, mark the actual distribution positions of each sampleable region in the image, group the distribution results by image frame, and each frame of the image forms a distribution record list. For example, the candidate blocks corresponding to frame T3 are C01, C03, and C05, and the corresponding positions are (120, 80), (160, 80), and (200, 120) respectively. Finally, combine and organize the spatial distribution data of all sampleable candidate blocks in all image frames to establish the distribution information of the sampleable cultivated land image blocks for subsequent multi-frame data fusion or partition sampling scheduling.

[0140] Please refer to Figure 2 , a remote sensing monitoring system for cultivated land protection, including:

[0141] The temporal gray-scale change extraction module obtains the remote sensing image sequence of the cultivated land area at different time nodes, extracts and sorts the gray-scale values of the pixels of the same piece of land, calculates the gray-scale difference between adjacent time nodes, and makes a ratio judgment on the difference in combination with the gray-scale mean value within the image frame, marks the pixels that exceed the gray-scale fluctuation reference value in continuous multiple changes, and generates a pixel region set;

[0142] The high-frequency change area screening module calls the pixel region set, sets the frequency threshold and marks the high-frequency areas, checks the spatial positions and coverage relationships of the marked areas, adjusts the sampling levels of the sampling dense areas, records the adjusted pixel sampling sorting, and generates a reference value for the temporal image sampling priority;

[0143] The albedo fluctuation screening module calls the reference value for the temporal image sampling priority, extracts the multi-band albedo sequence of the adjusted sampling area, calculates the albedo change rate within the time window, screens out the areas where the rate difference exceeds the albedo jump threshold, and obtains a set of sampling retention area segments;

[0144] The slope occlusion exclusion module calls the set of sampled reserved area segments, extracts the solar incidence angle and terrain slope values of the corresponding areas, calculates the solar direction projection length, judges the occlusion conditions, combines the regional reflection intensity variation range and the projection value, screens the areas exceeding the occlusion determination threshold, and obtains the set of occlusion exclusion image positions;

[0145] The cultivated land image block extraction module calls the set of occlusion exclusion image positions, marks the spatial distribution of the remaining areas, tracks the plot boundaries in the image frame, judges the boundary integrity and coherence, screens the areas that meet the sampling structure continuity, outputs the spatial index and position set, and generates the distributive information of the cultivable land image blocks that can be sampled.

[0146] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A remote sensing monitoring method for cultivated land protection, characterized in that, The following steps are involved: S1: Obtain remote sensing image sequences of cultivated land areas at differentiated time nodes, extract pixel grayscale values ​​of the same plot and arrange them by time, calculate the ratio of the difference between adjacent frames to the grayscale mean within the frame, mark the positions that continuously exceed the fluctuation reference ratio, and generate a pixel area set; S2: calling the pixel region set, setting the interference frequency threshold to filter the frequent regions, checking the position and coverage relationship frame by frame, lowering the priority of pixels in the densely sampled regions, and generating a time-series image sampling priority reference value; S3: calling the time-series image sampling priority reference value, extracting the multi-band albedo sequence of the corresponding area, setting the time window and calculating the inter-frame change rate, comparing the current frame change rate with the fluctuation value in the window, and obtaining the sampling retention area segment set; S4: calling the sampled reserved area fragment set, extracting the sun incidence angle and slope parameters, calculating the illumination projection length and combining it with the reflection intensity change amplitude, screening the area where the combined value exceeds the occlusion threshold, and obtaining the occlusion exclusion image position set; S5: calling the mask exclusion image position set, marking the remaining area, tracking the extension status of the plot boundary, judging the boundary integrity and coherence, and generating the sampleable cultivated land image block distribution information.

2. The remote sensing monitoring method for cultivated land protection according to claim 1, wherein The pixel area set includes pixel position, grayscale fluctuation ratio, and change continuity mark; the time series image sampling priority reference value includes sampling area, sampling frequency threshold, and pixel sampling order; the sampling retention area fragment set includes albedo sequence, fluctuation rate, and jump threshold; the occlusion exclusion image position set includes solar incidence angle, terrain slope, projection length, and occlusion judgment threshold; the sampleable cultivated land image block distribution information includes spatial index, position set, and sampling coherent area.

3. The remote sensing monitoring method for cultivated land protection according to claim 1, wherein The specific steps of S1 are: S101: obtaining a remote sensing image sequence of a cultivated land area at different time nodes, extracting and sorting the pixel grayscale values ​​of the same area, calculating the grayscale difference between adjacent time nodes, combining the grayscale mean value in the image frame, performing ratio calculation on the grayscale difference of the pixel points, and obtaining a pixel grayscale ratio sequence; S102: Based on the pixel grayscale ratio sequence, pixel positions whose grayscale ratio changes exceed a set reference ratio at multiple consecutive time nodes are screened, and grayscale fluctuation marked pixel point sets are obtained after aggregation; S103: calling the grayscale fluctuation marked pixel point set, analyzing the spatial distribution characteristics of adjacent pixels, merging high fluctuation pixels, and forming a grayscale change region set.

4. The remote sensing monitoring method for cultivated land protection according to claim 1, wherein, The grayscale difference calculation formula of adjacent time nodes is specifically as follows: Among them, represents the gray-scale difference of the pixel point (x, y) between the i-th and the (i + 1)-th frame images, represents the gray-scale value of the pixel point (x, y) in the i-th frame, μ i represents the average value of the gray-scale values of all pixel points in the i-th frame image, N represents the total number of pixels in the image, represents the algebraic sum of the gray-scale value changes of all pixel points in the image between the i-th and the (i + 1)-th frames.

5. The remote sensing monitoring method for cultivated land protection according to claim 1, characterized in that The specific steps of S2 are: S201: calling the grayscale change region set, setting a pixel frequency threshold, marking the region exceeding the threshold, checking the spatial position of the marked region by frame, analyzing the coverage, setting a threshold and screening the region with a higher frequency, and obtaining the distribution of high-frequency marked regions; S202: Based on the distribution of the high-frequency annotated area, select a location in the densely sampled area to perform a sampling down-adjustment operation, while retaining the original sampling data level of the unannotated area to obtain an adjusted pixel sampling level; S203: Invoke the adjusted pixel sampling level, record the sampling sorting structure of pixel points, organize them in chronological order, and establish the chronological image sampling priority result.

6. The remote sensing monitoring method for cultivated land protection according to claim 1, wherein The specific steps of S3 are as follows: S301: Invoke the chronological image sampling priority result, extract the multi-band albedo sequence corresponding to the adjusted sampling area, set a time window and calculate the albedo change rate between consecutive frames, organize the change rates for time nodes, and establish an albedo change rate sequence; S302: Based on the albedo change rate sequence, compare the difference between the change rate of the area in the current sampling frame and the fluctuation rate within the time window, and filter out the areas where the difference exceeds the albedo jump threshold to obtain the albedo stable area; S303: Invoke the albedo stable area, extract the spatial distribution information of the retained area, organize the area segment data in time series, and generate a sampling retained area segment set.

7. The remote sensing monitoring method for cultivated land protection according to claim 1, characterized in that The specific calculation formula for the albedo change rate value is as follows: Among them, represents the albedo change rate value of the pixel point (x, y) in the i-th frame image in band b, represents the albedo value of the pixel point (x, y) in band b in the i-th frame image, represents the average albedo value of all pixel points in band b in this frame image, represents the albedo value of the pixel point (x, y) in band b in the selected reference time frame, Δt ref,i represents the time interval between the current image frame and the reference image frame.

8. The remote sensing monitoring method for cultivated land protection according to claim 1, wherein The specific steps of S4 are as follows: S401: Invoke the sampling retained area segment set, extract the solar incidence angle and terrain slope values of the corresponding area, calculate the projection length in the solar direction based on the incidence angle and slope values, identify the illumination projection range of the area, and determine whether there is a slope occlusion condition to obtain the slope occlusion judgment result; S402: Based on the slope occlusion judgment result, extract the change amplitude of the reflection intensity of the area and perform a combined operation with the projection value to calculate the combined value of the area and obtain the occlusion influence combined value; S403: Invoke the occlusion influence combined value, filter out the areas where the combined value exceeds the occlusion determination threshold, extract the corresponding image positions and organize the spatial information to generate a shielding exclusion image position set.

9. The remote sensing monitoring method for cultivated land protection according to claim 1, characterized in that, The specific steps of S5 are as follows: S501: Invoke the shielding exclusion image position set, mark the distribution state of the remaining areas in the spatial grid, track the extended form of the corresponding plot boundary in the image frame, analyze the spatial connection of the areas based on the continuity characteristics of the adjacent areas of the boundary, and judge the integrity and coherence of the boundary to obtain the boundary coherent area; S502: Based on the boundary coherent area, filter out the areas that meet the continuous determination conditions of the sampling structure, and classify the qualified areas as candidate blocks to obtain the samplable candidate areas; S503: Invoke the samplable candidate areas, organize the spatial index information, and match the corresponding image positions to establish the distribution information of the samplable cultivated land image blocks.

10. A remote sensing monitoring system for cultivated land protection, characterized in that, According to a remote sensing monitoring method for cultivated land protection according to any one of claims 1-9, the system includes: The time series gray-scale change extraction module obtains the remote sensing image sequence of the cultivated land area at different time nodes, extracts and sorts the gray-scale values of the pixels of the same plot, calculates the gray-scale difference between adjacent time nodes, makes a ratio judgment on the difference in combination with the gray-scale mean value within the image frame, marks the pixels that exceed the gray-scale fluctuation reference value in continuous multiple changes, and generates a pixel area set; The high-frequency change area screening module calls the pixel area set, sets the frequency threshold and marks the high-frequency areas, checks the spatial position and coverage relationship of the marked areas, adjusts the sampling level of the sampling dense areas, records the adjusted pixel sampling order, and generates a reference value for the sampling priority of the time-series image; The albedo fluctuation screening module calls the reference value for the sampling priority of the time-series image, extracts the multi-band albedo sequence of the adjusted sampling area, calculates the albedo change rate within the time window, screens out the areas where the rate difference exceeds the albedo jump threshold, and obtains the set of sampling retention area segments; The slope occlusion exclusion module calls the set of sampling retention area segments, extracts the solar incidence angle and terrain slope value of the corresponding area, calculates the solar direction projection length, judges the occlusion condition, combines the regional reflection intensity change amplitude and the projection value, and screens out the areas that exceed the occlusion determination threshold to obtain the set of positions of the occlusion exclusion image; The cultivated land image block extraction module calls the set of positions of the occlusion exclusion image, marks the spatial distribution of the remaining areas, tracks the plot boundaries in the image frame, judges the integrity and coherence of the boundaries, screens out the areas that meet the sampling structure continuity, outputs the spatial index and position set, and generates the distribution information of the sampleable cultivated land image blocks.

Images