Method for Monitoring Grassland Vegetation Coverage Based on Continuous Photography

Abstract

A method for monitoring the grassland vegetation coverage based on continuous photography, comprising: determining an observation quadrat and continuously taking multiple vegetation photos of the vegetation within the observation quadrat throughout the entire growing season; cropping and geometrically correcting the multiple vegetation photos; obtaining the vegetation pixel ratio corresponding to the vegetation photos and comparing the vegetation pixel ratio with a preset second threshold; when the vegetation pixel ratio is greater than the second threshold, calculating the vegetation coverage corresponding to the vegetation photos by using the Otsu algorithm; otherwise, using the vegetation pixel ratio as the vegetation coverage; obtaining a dynamic change curve of the vegetation coverage based on the vegetation coverage corresponding to each vegetation photo. The present invention effectively improves the monitoring accuracy and efficiency of the dynamic change characteristics of the grassland vegetation coverage by combining continuous photography and the Otsu algorithm.

Description

Technical Field

[0001] The present invention belongs to the technical field of monitoring, and particularly relates to a method for monitoring grassland vegetation coverage based on continuous photography. Background Art

[0002] As an important ecological index of grassland ecosystems, vegetation coverage information can be used to reveal key ecological processes such as vegetation productivity, soil and water conservation ability, and phenological changes. However, in previous work, the monitoring methods of vegetation coverage mainly included manual visual inspection and remote sensing monitoring. The monitoring results of manual visual inspection highly depend on the personal experience of the measurer, and also result in a large amount of time and effort being consumed for observing the coverage dynamic changes at the plot level; while remote sensing monitoring has the problem that the spatio-temporal accuracy is too rough to show the details of coverage changes. Handheld digital cameras, as a means of near-surface remote sensing, are helpful for extracting vegetation coverage information. In previous work on monitoring vegetation using digital cameras, the camera would be fixed on a small house or iron tower at a certain distance from the monitoring point. However, the non-vertical angle of the fixed camera cannot accurately measure the vegetation coverage, is not easy to be used for monitoring with many individual small quadrats, and it is difficult to find a suitable fixed position on open grasslands, and there is a risk of being damaged by strong winds, animals, and residents.

[0003] Therefore, there is a need in the art for a new method for monitoring grassland vegetation coverage to solve the above problems. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for monitoring grassland vegetation coverage based on continuous photography, aiming to solve the problem of how to improve the monitoring accuracy and efficiency of grassland vegetation coverage.

[0005] To achieve the above purpose, the present invention adopts the following technical solutions:

[0006] A method for monitoring grassland vegetation coverage based on continuous photography, the method includes:

[0007] Determine an observation quadrat and continuously take multiple vegetation photos of the vegetation within the observation quadrat throughout the growing season;

[0008] Crop and geometrically correct the multiple vegetation photos;

[0009] Obtain the vegetation pixel ratio corresponding to the vegetation photo, wherein the vegetation pixel ratio is the ratio of vegetation pixels to the total number of pixels of green, red, and blue, the vegetation pixel is a green pixel whose green pixel value is greater than a preset first threshold, and the green pixel is a pixel whose green pixel value is greater than the red pixel value and greater than the blue pixel value;

[0010] Compare the vegetation pixel ratio with a preset second threshold;

[0011] When the vegetation pixel ratio is greater than the second threshold, the Otsu algorithm is used to calculate the vegetation coverage corresponding to the vegetation photo;

[0012] Otherwise, the vegetation pixel ratio is used as the vegetation coverage;

[0013] Based on the vegetation coverage corresponding to each vegetation photo, a dynamic change curve of the vegetation coverage is obtained.

[0014] Preferably, the step of "obtaining a dynamic change curve of the vegetation coverage based on the vegetation coverage corresponding to each vegetation photo" includes:

[0015] Based on the discrete data of the vegetation coverage corresponding to each vegetation photo, smoothing is performed using a double logistic curve equation or spline fitting to obtain a smooth dynamic change curve of the vegetation coverage.

[0016] Preferably, the size of the observation quadrat is 1m×1m, or not less than 0.5m×0.5m;

[0017] The observation quadrat defines the boundary of the photography area using a quadrat frame made of PVC or wood;

[0018] The effective pixels of the vegetation photo are not less than 10 million;

[0019] The height of the camera for taking the vegetation photo from the ground is 1m.

[0020] Preferably, the step of "cropping and geometrically correcting the multiple vegetation photos" includes:

[0021] The vegetation photo is sheared to remove the part outside the observation quadrat, and geometric correction is performed to obtain vegetation photos of a unified size.

[0022] The advantages of the present invention are as follows:

[0023] The method for monitoring the grassland vegetation coverage based on continuous photography provided by the present invention effectively improves the monitoring accuracy and efficiency of the dynamic change characteristics of the grassland vegetation coverage by combining continuous photography and the Otsu algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 is a schematic diagram of the main process of a method for monitoring the grassland vegetation coverage based on continuous photography according to the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0025] The present invention will be further described in detail below with reference to the drawings and embodiments.

[0026] The inventor's previous research has proven that digital photos are a convenient and fast method for monitoring the dynamic changes in grassland vegetation coverage, especially in field experiments with multiple fixed small quadrats, which can effectively save the time for plot surveys. After obtaining the photos, semi-automatic analysis is adopted, and a large number of pictures can be easily and quickly processed in a short time using computer software. In addition, the vegetation coverage obtained by the present invention is an objectively quantified index parameter, independent of the subjective judgment of observers, and has high data quality and comparability. The following will describe in detail the method for monitoring grassland vegetation coverage based on continuous photography provided by the present invention with reference to the accompanying drawings.

[0027] Refer to the attached Figure 1 , Figure 1 which exemplarily shows the main process of a method for monitoring grassland vegetation coverage based on continuous photography. As Figure 1 shown, the method for monitoring grassland vegetation coverage based on continuous photography provided in this embodiment includes:

[0028] Step S1: Determine the observation quadrat and continuously take multiple vegetation photos of the vegetation within the observation quadrat throughout the entire growing season.

[0029] Specifically, the size of the observation quadrat is preferably 1m × 1m, or not less than 0.5m × 0.5m. When taking photos for sampling, a quadrat frame made of PVC or wood (white color is the best) is needed to define the boundary of the photography area. When taking photos, it is recommended to use a Nikon D5100 digital camera with an effective pixel of not less than 10 million to ensure that the details of the vegetation information are retained. The height of the camera from the ground is 1m, keeping the camera horizontal and the lens facing the quadrat frame for shooting.

[0030] The dynamic monitoring of vegetation coverage usually includes the entire growing season, and the photo-taking time starts from the green-up period to the withering period. For example, in the grassland area of northern Hebei, the suitable time is from early May to early September, and the best effect is achieved when the photo-taking interval is less than 7 days.

[0031] Step S2: Crop and geometrically correct the multiple vegetation photos.

[0032] Specifically, after collecting the vegetation photos, the Photoshop software can be used to cut and geometrically correct the images. That is, remove the part outside the quadrat frame and only retain the vegetation part, making all the photos into squares of exactly the same size to obtain the final corrected sequence of vegetation photos. In this embodiment, the quadrat frame corresponding to each vegetation photo is 1m × 1m.

[0033] Step S3: Obtain the vegetation pixel ratio corresponding to the vegetation photos.

[0034] Specifically, the vegetation pixel ratio is the ratio of vegetation pixels to the total number of pixels of green, red, and blue. Vegetation pixels are green pixels with a green pixel value greater than a preset first threshold. Green pixels are pixels with a green pixel value greater than the red pixel value and greater than the blue pixel value.

[0035] In this embodiment, the first threshold is 30, that is, vegetation pixels are defined as green pixels with G > 30 (RGB value, maximum 255) among green pixels. Because when G ≤ 30, the pixel is close to white and is not determined as vegetation. The vegetation pixel ratio is the ratio of vegetation pixels to the total number of pixels of green, red, and blue, as shown in the following formula:

[0036] f V = N V / (N G + N R + N B )

[0037] In the formula, f V is the vegetation pixel ratio, N G , N R , N B respectively represent the number of pixels of green, red, and blue, and N V is the number of vegetation pixels.

[0038] Step S4: Compare the vegetation pixel ratio with a preset second threshold.

[0039] After obtaining the vegetation pixel ratio, it is compared with the second threshold. When the vegetation pixel ratio is greater than the second threshold, go to step S5; otherwise, go to step S6. In this embodiment, the second threshold is 20%.

[0040] Step S5: Calculate the vegetation coverage corresponding to the vegetation photo using the Otsu algorithm.

[0041] Specifically, this embodiment uses the Otsu algorithm to extract the coverage information of the vegetation photo. The Otsu algorithm is a reliable algorithm for distinguishing foreground from background and can be used for image binarization, and can be calculated in MATLAB software.

[0042] Step S6: Use the vegetation pixel ratio as the vegetation coverage.

[0043] Specifically, use the vegetation pixel ratio corresponding to the vegetation photo as the vegetation coverage corresponding to the vegetation photo.

[0044] Step S7: Obtain the dynamic change curve of vegetation coverage based on the vegetation coverage corresponding to each vegetation photo.

[0045] Specifically, based on the discrete data of the vegetation coverage corresponding to each vegetation photo, smoothing is performed using the double-logistic curve equation or spline fitting to obtain a smooth dynamic change curve of the vegetation coverage.

[0046] The above is the preferred embodiment of the present invention and the technical principles applied thereby. For those skilled in the art, any obvious changes such as equivalent transformations and simple substitutions based on the technical solution of the present invention without departing from the spirit and scope of the present invention shall fall within the protection scope of the present invention.

Claims

1. A method for monitoring the grassland vegetation coverage based on continuous photography, characterized in that, The method includes: Determining an observation quadrat and continuously taking multiple vegetation photos of the vegetation within the observation quadrat throughout the entire growing season; Cropping and geometrically correcting the multiple vegetation photos; Obtaining the vegetation pixel ratio corresponding to the vegetation photos, where the vegetation pixel ratio is the ratio of vegetation pixels to the total number of pixels of green, red, and blue, and the vegetation pixels are green pixels with a green pixel value greater than a preset first threshold, and the green pixels are pixels with a green pixel value greater than the red pixel value and greater than the blue pixel value; Comparing the vegetation pixel ratio with a preset second threshold; When the vegetation pixel ratio is greater than the second threshold, calculating the vegetation coverage corresponding to the vegetation photo using the great law algorithm; Otherwise, using the vegetation pixel ratio as the vegetation coverage; Obtaining a dynamic change curve of vegetation coverage based on the vegetation coverage corresponding to each vegetation photo.

2. The grassland vegetation coverage monitoring method based on continuous photography according to claim 1, wherein The step of "obtaining a dynamic change curve of vegetation coverage based on the vegetation coverage corresponding to each vegetation photo" includes: Based on the discrete data of the vegetation coverage corresponding to each vegetation photo, performing smoothing using a double logistic curve equation or spline fitting to obtain a smooth dynamic change curve of vegetation coverage.

3. The method for monitoring grassland vegetation coverage based on continuous photography according to claim 2, wherein: The size of the observation quadrat is 1m×1m, or not less than 0.5m×0.5m; The observation quadrat defines the boundary of the photography area using a quadrat frame made of PVC or wood; The effective pixels of the vegetation photo are not less than 10 million; The height of the camera for taking the vegetation photo from the ground is 1m.

4. The method for monitoring grassland vegetation coverage based on continuous photography according to claim 3, wherein, The step of "cropping and geometrically correcting the multiple vegetation photos" includes: Shearing the vegetation photo to remove the part outside the observation quadrat, and performing geometric correction to obtain vegetation photos of a unified size.

Images