Shadow extraction method facing ecological environment parameter remote sensing inversion

Abstract

The invention discloses a shadow extraction method facing ecological environment parameter remote sensing inversion. The method comprises the following steps of performing radiometric calibration andatmospheric radiation correction on an original remote sensing image to obtain a remote sensing image result map; calculating a vegetation index, and calculating a shadow index according to the obtained vegetation index so as to obtain a shadow index information space distribution map; performing enhanced algorithm treatment of an image on the remote sensing image result map, and collecting the picture element sample of a shadow as a sample area at random; statistically calculating cumulative frequency and statistical variable, which change from small to big, of a shadow index number value corresponding to the sample area, and setting up the range of a threshold; according to the obtained range of the threshold, making and extracting a two-value mask pattern of shadow information, enablingthe made two-value mask pattern and the remote sensing image result map to be subjected to space superposition, and performing wave band product calculation to quickly extract an accurate space distribution map of shadow information in a remote sensing image. The shadow extraction method is simple and flexible in operational process and easy to popularize and apply, and improves the accuracy of ecological environment parameter information remote sensing inversion.

Description

technical field [0001] The invention relates to a shadow extraction method for remote sensing inversion of ecological environment parameter information, belonging to the technical field of ecological environment quantitative monitoring and evaluation. Background technique [0002] The ecological environment is the foundation of the sustainable development of the social economy, and the quantitative evaluation of the ecological environment is an important basis for implementing the regional sustainable development strategy. Ecological environment parameters are biological, physical, and chemical parameters used to describe the ecology and environment of the earth's surface, and can quantitatively reflect the status of the regional ecological environment. The measurement density of traditional ecological environment parameter information data is generally small, and the time interval is often very long, lacking continuity and consistency in time and space; at the same time, th...

AI Technical Summary

Problems solved by technology

The first type of method needs information such as the geometric shape of the occluded object, the sun azimuth, sensor parameters, etc., so as to establish a corresponding geometric model to realize the extraction of shadow information, but in practical applications, a lot of information is not easy to obtain, and has great limitations. sex

The second type of method mainly uses the spectrum, texture, color and edge properties of the shadow, combined with certain algorithms, to realize the extraction of shadow information. However, due to its own constraints, the existing methods are only in some specific fields. or specific objects have achieved better results

For example, the detection method based on color invariance has many restrictions on the scene, assuming that the shadow area is a non-textured flat surface, which must be a single light; the histogram-based method has the defects of strong reflection terrain, ground objects and water bodies. More importantly, it is only suitable for gray-scale images; many studies have proposed methods such as homomorphic filtering based on the fact that the brightness of shadowed areas is lower than that of non-shaded areas in HIS space, but these methods are easy to convert water bodies, Low-brightness features were misjudged as shadows; Wei Minglu et al. combined HSV transformation and the principle of region growth, and used the double threshold method to propose an improved shadow detection method, which has a better ability to identify shadow areas, but the region growth The result depends on the selection of the initial seed point, the randomness is large, and the stability of the method is poor

Combining water body index and HIS transformation, Luo Yan et al. can effectively remove water body information, but for different remote sensing images, it is necessary to refer to the OIF algorithm to find the best band combination and then perform HIS transformation, which has poor universality.

In recent years, combined with the nature of shadows, some object-oriented shadow extraction methods have been developed, which have a better recognition effect on shadow areas, but such methods are also aimed at specific areas, and are more common in the extraction of building shadows in small urban areas, and the extraction The process is complex and cumbersome, and the universality is not strong. Sometimes it is necessary to adjust the relevant constants according to the influence of the illumination, the proportion of the object type, and the sensor.

Images