Interpretation of aerial photos is a process of extracting the required thematic information from remote sensing images and other related materials according to the classification system. When interpreting, we should not only start with the direct characteristics of land use categories, such as color, geometric pattern, shape and size of special images, but also combine the indirect signs reflected by the distribution characteristics and relationships of different categories in topography and geomorphology, and also consider the notes on the same scale topographic map for logical reasoning and comprehensive analysis, and conduct field visits and interviews when necessary to finally establish the interpretation signs. For example, the interpretation signs of various land use types in this mining area are as follows.
14 1. Gully dry land: it is an important cultivated land in the valley. Under the stereoscopic lens, it can be observed that it is located in the valley, including gully and terrace, and the plot is smaller than the flat dry land. Its images are light yellow, goose yellow and light cyan with irregular edges.
142. Terraced dryland: Generally, it is constructed manually with uniform texture. There are obvious steps in the stereo. There are yellow spots and dark red weeds or shrubs at the junction of the upper and lower steps. Ridges are similar to regular circular arc strips arranged in concentric circles, with light yellow and light green colors, which are distributed in mountainous areas and areas with large slope changes.
143. Sloping dry land: The slope can be observed under stereo microscope, and there is no obvious ridge. Smooth and irregular texture, delicate texture and great contrast. Generally, the sunny slope is light yellow and white, and the shady slope is light blue and green.
144. Sichuan dry land: the terrain is relatively flat and open, with a large area. The image is regular and blocky, with high brightness and light yellow and light green tones. You can see white paths criss-crossing the fields.
15. Vegetable field: The image of vegetable field is clearly visible. The plot is small and rectangular, with a checkered carpet-like structure inside. The image is vermilion with a little noise. Mostly distributed in residential areas, ponds and ditches.
2 1. Orchard: The spacing between rows and plants is regular, and the periphery is clear, which is reflected as regular dark red dots and shadows in the image.
2 1 1. Apple Orchard: Red or magenta dots regularly distributed on the image.
2 12. Persimmon Garden: The plot is similar to dry land, with small area, crimson color and rough texture.
2 13. Other orchards: red or magenta, with clear stripes and obvious boundaries with surrounding areas.
22. Other gardens: sparsely dotted, with irregular stripes and dark red to red. The boundary with the surrounding area is obvious.
3 1. Woodland: The trees grow densely and the trunks are tall, showing a continuous large-scale pattern. Usually, the crowns are connected together in a purple strip. The boundaries are obvious but irregular, mostly distributed in mountainous areas and residential areas.
3 1 1. Coniferous forest: dark red, mottled texture, pointed crown, sharp shadow under stereomicroscope.
3 12. Broad-leaved forest: the crown of the tree is red, forming grape string texture, and the shadow is oval.
32. Shrub: Brown or reddish brown. The schlieren structure is rough, the pattern spots are irregular in shape and spherical under stereomicroscope.
33. Sparse forest: the image is reddish or lavender, with fine internal structure and irregular distribution.
36. Nursery: It is generally distributed on the edge of residential areas, with regular periphery, clear spacing between rows and plants, small crown with bright red stripes, rectangular or square plots and checkered interior.
4. Grassland: velvet structure, even and smooth shadow pattern, yellow-green, with red spots.
4 1. Natural grassland: light gray or reddish brown, with blurred boundaries, mostly distributed in mountainous areas.
42. Artificial grassland: regular shape, red, distributed in lowlands and around institutions.
5 1. Town: There are many streets on the land of the city. These streets are arranged regularly and obviously, and the color is dark blue. There are neat and dense residential areas, which are located next to the main traffic lines and have a large area.
52. Rural residential area: The image has irregular outline, dense houses, blue-gray color, curved internal roads, many trees inside, small area and rural roads.
53. Independent industrial and mining land: Generally, its boundary is surrounded by walls, and the internal houses are tall and tidy. Generally, there are yards and special railways or highways, and the greening conditions are good.
53 1. Land for industrial and mining facilities: blue-gray, with various forms and obvious location of large equipment.
532. Courtyard: The color is mostly cyan or bluish gray, showing a regular accumulation. You can see many ground objects (such as wood and steel pipes) and have certain facilities, such as overhead driving, lifting and transportation machinery, and transportation lanes in waste hills. , connected by railway or highway, and large mining fields are connected by special railway lines.
533. Industrial and mining residential land: In large factories and mines, houses are arranged in a regular pattern, mainly in blue-gray tone, mostly isolated from production areas or yards, and there are occasional dark red spots (trees) in front of and behind the houses.
534. Land for industrial, mining and public facilities: The outline is obvious and has a certain geometric shape, such as square or rectangle. Generally, the space is large, the houses are tall and tidy, and the surrounding greening is good.
535. Small independent industrial and mining land: As an independent whole, there are coal gangue hills or chimneys, as well as relatively small material yards, which are connected by simple roads. The image is obvious, and features such as chimney and kiln roof can be seen. There are few surrounding houses, generally small coal mines, iron works, brick and tile factories and other industrial enterprises.
55. Special land: This is a kind of land that is difficult to distinguish. Because of its small distribution range, it can be delineated by on-site mapping.
6 1. Railway land: the railway is cyan, narrow and long, streamlined, and the special railway line in the mining area goes straight to the mining area stockyard, and the arc is large and even when turning.
62. Highway land: the asphalt pavement is blue-gray with dark red roadside Woods; Cement road or gravel road is light gray and grayish white; The curvature of the road curve is large, and the vehicles driving on the road are obviously visible.
63. Rural roads: grayish white, reticulate, narrow and curved, connected with villages and fields.
7 1. River: The river is generally a curved streamlined black strip, and the bridge is located on it when the road passes. The riverbed edge is irregular and the river width is uneven, and there are usually vegetable fields nearby.
72. Reservoir: The water surface is black, with a large area and dams, mostly gourd-shaped.
73. Pits and ponds: Generally, there are regular boundaries, which are round, oval or rectangular. The water surface is mostly black, with a small area and easy to distinguish.
78. Hydraulic structures: white stripes, located downstream of the reservoir.
8 1. wasteland: gray or brownish red, irregularly distributed in a large area, with uniform texture and no three-dimensional effect.
86. Bare rock or gravel land: white, irregular in shape, with gray bands in the middle, mostly distributed in mountainous areas.
87. Ridge: Curved, linear, gray or brown with shading, located between sloping dry land and terraced dry land.
According to the above-mentioned image interpretation marks, each color exterior photo is analyzed and interpreted one by one, and the interpretation results can be drawn on the 0.05mm thick transparent polyester film covered on the color infrared aerial photo, so as to take samples on the spot and check the correct rate. At the same time, according to the mark, the boundary of the type points that cannot be directly interpreted is drawn on the image, and its category attributes are adjusted on the spot.
Verification of land use interpretation results in 3.3.2.2 mining area
In order to improve the accuracy of category attributes of land use maps, it is necessary to verify the interpretation results on the spot. The verification adopts sampling method, and the number of samples is determined according to the sampling theory and the number of patches explained.
(1) sampling ratio
First of all, it is necessary to determine how much error is allowed in the total area of land use types. Because in the area statistics of land use types, the error of category interpretation has the greatest influence on the error of classified area. Only by improving the interpretation accuracy can the accuracy of area statistics be guaranteed. According to the Technical Specification for Land Use Survey, the accuracy of map interpretation should reach over 95%. Therefore, take α = 0.05 and determine the number of random samples within 95% confidence level.
Assume that the pattern events of the sampled land types are normally distributed, and the number of samples is given by the following formula.
Study on environmental dynamic monitoring and analysis in industrial and mining areas
Where n is the number of samples to be sampled; P is the ratio of the number of correctly explained plaques to the total number of plaques in the population; Q is the ratio of the number of incorrectly explained patches to the total number of patches in the population; D is the allowable error; N is the total number of spots; T is a multiple of standard deviation.
Because the values of p and q are unknown, their estimated values p and q are used in actual sampling.
When the value of n is large, the above formula can be approximately expressed as:
Study on environmental dynamic monitoring and analysis in industrial and mining areas
When n0/N can be ignored, n=n0.
When n0/N is not negligible,
Study on environmental dynamic monitoring and analysis in industrial and mining areas
When the confidence is 95%, let t = 2, and P can be obtained according to the data of land use survey in mining area.
(2) Sampling inspection results
In sampling verification, the field category is compared with the interpretation category on the image, and the interpretation result matrix is counted, and the judgment rate is calculated accordingly.
Mapping method of land use map in 3.3.2.3
The drawing process is as follows:
(1) First, the aerial photos (1: 10000) are analyzed and interpreted in detail indoors, and the land use types are divided into patches. Then cover the aerial photos with polyester film, and draw the outline boundary of the type with a pencil in the use area. Then it is verified on the spot, focusing on the categories that are easy to interpret errors.
(2) Taking the topographic map as the working base map, and using the aerial photography plotter, the photos with different scales are accurately nested with the topographic map. Find out five obvious feature points (flat areas) corresponding to topographic map on the boundary of effective area of each interpretation slice, or find out control points (hilly terrain) in blocks, and then transfer land use map spots to topographic map through scale adjustment, instrument and base map movement. , forming a sketch of land use map.
(3) Put the sketch of land use together with the boundary and coordinates of topographic map on the digital board, with the support of GIS, complete the input, editing and drawing of graphics, automatically count the area of land use types, automatically draw the result map and make comprehensive analysis of various information.
With the help of various commands provided by GIS, the functions of graphic input, topological relationship generation, graphic editing, geometric transformation, graphic splicing, statistical calculation, automatic drawing and so on are carried out to complete the automatic drawing of the result map (sketch) and the statistics of land use area. The flow chart is shown in Figure 3-4.
Figure 3-4 Process of Drawing Land Use Map with GIS
Statistics of land use area in 3.3.2.4
The conventional statistical method of land area is generally to measure the area of each point on the land use map with orthogonal instruments, and then adjust it, which has a large workload and the measurement accuracy varies with different instruments and personnel. When using GIS to operate the attribute table, because the land use map uses the Gaussian coordinates of the actual section to correct each spot and match it with the section, its error has been automatically shared, and it is not necessary to adjust it again. The main reason that affects the area accuracy is the error in the process of digitization, which can be ignored. In addition, the general mining area includes multiple mines, so at the same time of digitization, the mine field boundary of each mine is input, so as to count the land use area of each mine.