This data is the vegetation coverage data set in a growth cycle of farmland, wetland, Gobi, desert and desert observation in Yingke oasis. Data observation starts from May 25, 2012 to September 14, 2012. It is observed once every 5 days before late July and once every 10 days thereafter. Measuring instrument and principle: The vegetation coverage of farmland, wetland, Gobi, desert and desert in Yingke oasis was measured by digital camera. The quadrat design, photo shooting method and data processing method have been analyzed and considered. It is described in several articles: 0 Measuring instrument: a simple observation frame is equipped with a digital camera. The digital camera is placed on the instrument platform at the front end of the support rod to keep the shooting vertical and downward, and remotely control the camera to measure data. The observation frame can be used to change the shooting height of the camera and realize targeted measurement for different types of vegetation. one Quadrat setting and "true value" acquisition: quadrat size of low vegetation such as corn 10 × 10m, fruit tree quadrat 30m × 30 meters. Take photos along two diagonals in turn during each measurement, and take a total of 9 photos (less than 9 when the surface coverage is very uniform), which are evenly distributed in the quadrat. After 9 photos are processed to obtain their respective coverage, take the average, and finally get the "true value" of the coverage of a quadrat. two Shooting method: for low vegetation such as corn, directly use the observation frame to ensure that the height of the camera on the observation frame from the vegetation canopy is much greater than the vegetation canopy, sample along the diagonal in the square quadrat, and then make arithmetic average. When the field of view angle is small (< 30 °), the field of view includes more than 2 full cycle ridges, and the side length of the photo is parallel to the ridges; For higher vegetation trees, take photos from bottom to top under the canopy, overlay and cooperate with the shooting of low vegetation on the ground under the canopy from top to bottom to obtain the coverage near the plants, and then take photos of low vegetation in the non canopy projection area between plants to calculate the coverage of plant gap. Finally, the average area of tree crown is obtained by crown projection method. According to the ridge distance, the area ratio between the plant crown and the plant gap is calculated, and the coverage of the whole quadrat is weighted. three Data processing method: an automatic classification method is adopted. For details, see article 3 of "references" (Liu et al., 2012). The RGB color space is converted to the lab space which is easier to distinguish the green vegetation, and the histogram of the green component A is clustered to separate the two components of green vegetation and non green background, so as to obtain the vegetation coverage of a single photo. The advantage of this method is that its algorithm is simple, easy to implement, and has high degree of automation and precision. In the future, more rapid, automatic and accurate classification methods are needed to give full play to the advantages of digital camera methods. Supporting data: The type, plant height, ridge width, row width and shooting height of vegetation are recorded in the record table, and the scene photos and ridge photos (farmland) taken by digital camera are attached. Data processing: Based on the classification method in the digital image, the vegetation coverage of the photo representative quadrat is obtained after classifying the vegetation and non vegetation pixels.
| collect time | 2012/05/25 - 2012/09/14 |
|---|---|
| collect place | Heihe River Basin, pilot area of artificial oasis in the middle reaches, desert, Gobi, wetland, desert, farmland |
| data size | 8.9 GiB |
| data format | Excel |
| Coordinate system | WGS84 |
The simple observation frame is equipped with a digital camera. The digital camera is placed on the instrument platform at the front end of the support rod to keep the shooting vertical and downward, and remotely control the camera to measure data. Quadrat setting and "true value" acquisition: quadrat size of low vegetation such as corn 10 × 10m, fruit tree quadrat 30m × 30 meters. Take photos along two diagonals in turn during each measurement, and take a total of 9 photos (less than 9 when the surface coverage is very uniform), which are evenly distributed in the quadrat. After 9 photos are processed to obtain their respective coverage, take the average, and finally get the "true value" of the coverage of a quadrat
Shooting method: for low vegetation such as corn, directly use the observation frame to ensure that the height of the camera on the observation frame from the vegetation canopy is much greater than the vegetation canopy, sample along the diagonal in the square quadrat, and then make arithmetic average. Small angle in field of view (In the case of< 30 °), the field of view includes more than 2 full cycle ridges, and the side length of the photo is parallel to the ridges; For higher vegetation trees, take photos from bottom to top under the canopy, overlay and cooperate with the shooting of low vegetation on the ground under the canopy from top to bottom to obtain the coverage near the plants, and then take photos of low vegetation in the non canopy projection area between plants to calculate the coverage of plant gap. Finally, the average area of tree crown is obtained by crown projection method. According to the ridge distance, the area ratio between the plant crown and the plant gap is calculated, and the coverage of the whole quadrat is weighted.
Data processing method: an automatic classification method is adopted. See article 0 of 5 "recommended references" for details. The RGB color space is converted to the lab space which is easier to distinguish the green vegetation, and the histogram of the green component A is clustered to separate the two components of green vegetation and non green background, so as to obtain the vegetation coverage of a single photo. The advantage of this method is that its algorithm is simple, easy to implement, and has high degree of automation and precision. In the future, more rapid, automatic and accurate classification methods are needed to give full play to the advantages of digital camera methods
Good data quality
This work is licensed under a
Creative
Commons Attribution 4.0 International License.
| # | title | file size |
|---|---|---|
| 1 | _ncdc_meta_.json | 8.9 KiB |
| 2 | 黑河生态水文遥感试验:黑河流域中游植被覆盖度数据集(2012.05.25-09.14) |
Vegetation airborne ground remote sensing fisheye camera vegetation coverage
gobi Desert Artificial oasis experimental area in the middle reaches desert Wetlands farmland Heihe River Basin
©Copyright 2005-. Northwest Institute of Eco-Environment and Resources, CAS.
Donggang West Road 320, Lanzhou, Gansu, China (730000)

