Chang Guang Satellite Technology Co., Ltd.

Remote Sensing Satellite Image Processing And Its Development Trend

When the remote sensing satellite in the air acquires the digital image of the earth and sends it back to the ground, is the work over? The answer is obviously no, on the contrary, this is the beginning of remote sensing digital image processing.

Remote sensing digital image (hereinafter referred to as "remote sensing image") is a digital remote sensing image. Different regions and objects on the earth surface can reflect or radiate electromagnetic waves of different wavelengths. Using this feature, remote sensing systems can produce different remote sensing digital images.

It is the imaging range and fineness of remote sensing images that make them stand apart from ordinary digital images, that is, electronic photos taken at ordinary times. The photography area of remote sensing satellite is a global macro dimension. Each pixel in the image corresponds to a certain number, a certain number or part of ground objects in the three-dimensional real world. According to the different resolution of satellite imaging, one of the pixels may be a tree, a car or a window of a building.

Therefore, the brightness value (DN value, Digital Number) of each pixel in the image has important information significance. To obtain accurate information, users need to manage, convert, correct, enhance, and extract a series of "magic operations" for pixels in satellite images according to their own application goals, so as to facilitate the subsequent in-depth mining and business integration applications.

DN value (Digital Number): the pixel brightness value of remote sensing image, recording the gray value of ground objects. It has no unit and is an integer value. The value is related to the sensor's radiation resolution, surface emissivity, atmospheric transmittance and scattering rate, and reflects the surface radiance.

Today, let's find out what "divine operations" are and how to apply them? And in today's rapid development of remote sensing industry, will the impact of high-frequency data output, algorithms and artificial intelligence change the traditional mode and underlying logic of "divine operation"?

1. What is remote sensing image processing?

Remote sensing image processing is a process of using computer image processing system to perform a series of operations on pixels in remote sensing images.

Remote sensing images contain a lot of information, which can be effectively analyzed and extracted only after digitization (sampling and quantization of imaging system, digital storage). On this basis, the image data processing "reprocessing", such as correcting the graph alignment coordinates and enhancing the contour of the ground objects, can greatly improve the accuracy of image processing and the efficiency of information extraction. This process can be called "remote sensing digital image processing".

As a basic and important part of the process of "Earth observation", in the satellite application industry chain, the remote sensing image processing link is in an important position in the middle and lower reaches, connecting the past and the future. The front end undertakes satellite ground facilities, and the back end provides "ready" data services or tools for specific business applications in agriculture, forestry, meteorology, natural resources and other industries.

2. Why is remote sensing image processing the "only way" for applications?

When we see neat and beautiful digital earth products such as Google Earth, or thematic maps or interpretation maps of remote sensing satellite applications in natural resource management, environmental protection, agriculture, meteorology and other fields, they need to go through the middle "baptism" of image processing.

Because remote sensing satellites "work" at high altitude, their imaging environment is far more complex than our daily ground photography environment, and they will encounter geometric deformation, distortion, blurring, noise, etc. caused by system and non system factors such as the instability of sensors, curvature of the earth, atmospheric conditions, illumination changes, terrain changes, etc. The remote sensing data center conducts preliminary processing such as strip removal and geometric coarse correction on the image. When the data reaches each end user, it needs to do further fine processing on the data to make it more close to the real world physical space environment and coordinates. It also conducts professional processing according to its own business analysis objectives to prepare for the next remote sensing image analysis, interpretation and business application.

In general, the main objectives of remote sensing image processing are as follows:

Image correction: restore and restore images. Before information extraction, the remote sensing image must be corrected, so that the image can correctly reflect the actual surface feature information or physical process.

Image enhancement: suppress or remove image noise. In order to make the surface feature information contained in the remote sensing image more readable, and the objects of interest more prominent, easy to understand and interpret, it is necessary to enhance the overall image or specific surface feature information.

Information extraction: according to the spectral and geometric characteristics of the ground objects, determine the extraction rules of different ground object information. On this basis, use the rules to extract various useful ground object information from the corrected remote sensing data.

3. How is remote sensing data processing changing?

Remote sensing data processing is more like the "raw material rough processing" link in production and manufacturing, and is also a precursor to the intelligent application and business integration of remote sensing image data. From the previous introduction, the process is also more complex and professional.

As an important part of the industrialization of earth observation and remote sensing, remote sensing data processing in the middle and lower reaches of the industry has also been impacted by the era of big data. It is responding to this trend and changing, moving towards real-time, standardization, scale and automation.

In the digital transformation of enterprises, it is often said that all traditional industries are worth doing again with digitalization, as are traditional data production and information service industries, and their models and processes are worth doing again with algorithms and AI.

When algorithms and artificial intelligence gradually penetrate the remote sensing data processing link, they can solve many problems in the remote sensing industry data production service, such as long data distribution cycle and link, many processing links, accuracy and consistency of massive data processing, which we can regard as "automated batch processing".

When the midstream algorithm engine solves the problems of data service, data calculation efficiency and automation process, more refined application data products suitable for various vertical subdivision scenes will appear in the downstream. In the remote sensing image information extraction link described above, with the participation of AI and algorithms, there will also be many efficient automation functions, such as target recognition, surface feature extraction, surface feature classification, change detection, etc, Gradually help people improve the efficiency of interpretation, and form a "intelligent information mining" mechanism in the downstream of remote sensing industry.

We can see that from the source of remote sensing data acquisition, to data processing, to terminal applications, its efficiency is inseparable from the underlying data model. With the trend of satellite Internet and Earth observation constellation gradually forming, only by standardizing the process of data acquisition, processing and sharing, can the large-scale, automated, and streamlined remote sensing industry better play a role in the digital transformation of government and enterprises, It also truly ushers in the era of space-time big data.