تحليل صور الاستشعار عن بعد.....Analysis of remote sensing images

Analysis of remote sensing images

One of the technical difficulties in the field of remote sensing in dealing with an output of this sensor, as the volume of information that is collected is so enormous, so it is nearly impossible to deal with it manually, hence it was necessary to develop programs for computers can deal with these photographs and radar , and transform them into information understandable and useful, and requires the analysis of satellite and aerial images and interpreted automatically, which is in the field of "Artificial Intelligence" Artificial Intelligence (AI).

First: "Apollo"

Uploaded on spacecraft "Apollo 9", which was launched in 1969, a group of cameras, taking photos of the Earth, a number of wavelengths, at the same time, these images showed that this technique, known as the "multi-spectral imaging" Multispectral Imaging could be used in several useful applications, such as discrimination between the vegetation healthy and infected pests, and in the work of accurate maps, and control of pollution and desertification, and so on.

II: the evolution of imaging techniques

Satellite images has led an important role in espionage operations during the Cold War era, and as a result of that event; significant evolution in imaging techniques, and how to deal with the satellite images. He started photography using film, which was returned to the land to Thmeidah, and then get bigger pictures.

This was followed by imaging using sensitive electronic circuits, where the information transmitted in the form of numbers, is collected automatically, using sophisticated computers, and then developed after the collection of information technology in the multi-spectra, which allows for night vision, or reading the difference in temperature between the body and another, and imaging radar, which penetrates the clouds.

III: Images of the space shuttle "Endeavor"

Reported the National Aeronautics and Space Administration "NASA" NASA that the pictures, drawn by the space shuttle "Endeavor", is better than those of existing images of the Earth, and most of them will remain confidential, and will be used exclusively for defense and intelligence of America.

The six astronauts aboard the shuttle had been extended pole, in space successfully, the first time held the structure in the vicinity of zero gravity. And the form of D Alsarip, which ended after six hours on a shuttle in orbit, one of the most important stages of this task, which aims to take a picture of the topography, the three-dimensional, the Earth's radar thanks to a very clear evolution.

Has allowed Alsarip capture two images of the Earth at the same time, using a technique known technology "overlapping packs radar waves." The picture was carried out by the first radar antennas in the shuttle, and the other by the antenna is installed in the party Alsarip.

The shuttle crew spent days in space, which put out radar signals to the cities, fields, mountains, forests, and almost anything that enters in the formation of the Earth's surface between the poles. In the echo signals into space, the antenna mounted on the pole for the shuttle collected images similar to movies, or holograms, to form a better picture of the topography of the land, obtained so far.

IV: Getting the images

It is known that the process of producing photographic images of the features by recording the reflection of light visible to the eye it, and satellite sensors, but the scanner does not "see" only the visible light, but "improved" with radiation in the electromagnetic spectrum, which is located behind the red and thus beyond the limited perception of the eye, which is called "the area near infra-red." Hence, the satellite's ability to "vision" multi-spectral, and the possibility to give extraordinary information.

Since different things, animal and plant life, minerals, reflect sunlight in different ways, both of them said: He has a "spectral fingerprint", Vahadjr sand reflects light in the infrared region is different from the reverse of the girl child, and differ from wet soil dry.

As different implications in maize than in wheat, and so on. Such fingerprints that they can be discriminated after careful analysis of the catalog information, and to address the computer on the ground. In this treatment, it becomes what is hidden necessarily visible through the use of colors "false" These photos provide a look unusual.

Fifth: modern methods of image capture and processing

In a test of the new method of looking at the Earth, highlighted the radar imaging system, which is carried by satellite, large sections of land, achieving impressive results. This led modern technologies, and electronic processing of data to a revolutionary change in the science of photography and art. Have participated in bringing about this revolution of satellites and computers, which made the new images, more able to understand the world. And perhaps more able to manage it, The images seek to simplify the world to the standards allow his understanding of the human person.

And the degree of differentiation Resolution imaging systems in the modern satellites can accurately notes the changes in the intensity of light reflected from the Earth's surface. The sense that each sensor responds to changes in light intensity, and the results stored in digital form, each number representing an element of the picture, the picture elements are grouped by addressing the computer for image formation.

Sixth: templates images substantive

Use of modern satellite "images plotters objectivity". In addition to its analytical sensitivity and high-light the differences, these sensors could not "see" more, because it measures reflected light in six wavelengths; four of them in the IR, as well as to capture the thermal channel temperature surfaces.

It is intended for images that served a specific purpose, or to explain a particular topic, it may be the distribution of soil or vegetation, or flow ice, or snow cover, or geological faults, or patterns of land use. All of these are tools to make drawing pictures by satellite.

Has developed technology for obtaining remote sensing images, in terms of clarity and precision. And numerous means of photography, including:

1. The video

And is determined by the two cameras, the angle between the Adstehma small; increases the field of view of the eyeglasses, and during the flight of the moon is in the field of vision strip of the Earth's surface, a width of thousand kilometers or more, the resulting image is recorded on magnetic tape, and then transmitted to the ground, when passing Moon over one of the reception centers for information, and imaging camera can only open spaces, while impeded by darkness and clouds.

2. Imaging infrared

The infrared devices measure the thermal radiation emitted by the Earth's surface, and access to thermal footprint of a region can determine the type of micro-components of the image (see the table of the relationship between body temperature and type of infrared rays that radiate), and playing computer major role in determining form of image, (see the picture of the amount of water in the atmosphere).

3. Photography

This was done using the imaging light cameras, with a powerful lens, to get the fine details, this is photography day, in good weather conditions, and this is the best method to obtain information confirming the three-dimensional image.

4. X-ray imaging

Modern techniques is a method of remote sensing, and requires a huge types of cameras. These rays can penetrate the sites to varying degrees to determine what's inside, and uses this method in space stations or satellites huge, weighing more than tons.

Seventh: The components of the image

An image sensor mounted remotely from the small areas equal, called picture elements, "pixel" Pixel, be visible in the lines and columns regularly. And imagine that the image consists of a huge amount of these elements, each of which represents a floor space varies from one satellite to another, it is 57 × 79 m in the first generation of satellites, "Landsat" of America, and 10 × 10 m or
20 × 20 m in the case of the French satellite "SPOT".

Of this area is emitting electromagnetic or repercussions, picked up by satellite devices, which send them to ground receiving stations. The number and the digital data to any component on the number of channels, which represent the wavelengths, which it receives spectral emissions or reflections.

VIII: image processing

Sher concept of image processing to the use of computers for data processing, stored in digital form. The aim of the processors in its entirety to further clarify the picture by increasing the contrast Contrast or divided into groups, or by other means, and are expressed in different colors, so as to reach the widest possible information about the objects in question.

Is a digital image processing of the most important techniques used in the field of remote sensing, and has helped in the application of this technique can be obtained on the data in digital form, and multi-wavelength on the one hand, and on the other hand witnessed the great development of computers, in terms of speed of processing of the data, the great potential to store. The key benefits of digital processing technology diversity, and the possibility of recurrence, and maintaining the accuracy of the data origin.

The computer linked to the pictures Baalaqtin basic pillars: image processing, and compilation of images by Computer Computer Graphics. The images are processed in a computerized three basic stages:

1. Insert the picture to be addressed to the memory of the computer.

2. Image processing.

3. Output image after treatment.

IX: image processing computer

Thus, the image processing computer-like, in principle, and stages, information processing computer. But is in fact the address of the visual information of the computer. Images, as information is visible, does not consist of separate units, Kalohrv and numbers, can be made of the computer, usually through input devices recognized, such as a keyboard, but that any picture that represents a wealth of ongoing information and extended in the framework.

On the other hand, the computer can only deal with the data, no Kalohrv and numbers, and in dealing with these data that converts each statement to a binary number Binary. This is called the image processing using computers, "digital image processing," Digital Image Processing.

Tenth: the introduction of image

Include the introduction of phase image to the computer memory mainly on the process of converting continuous flow of information, within the picture frame, into digital information sequence, this time dubbed the "survey and punctuation", where the image is divided into small elements called Pixel. In digital image processing, the image is divided into elements of equal sizes, here you can imagine that the image is made up of small squares.

The process of surveying and numbering give Alastdhaep intensity values for each of these elements. And described the value of each element refers to the intensity of illumination. In the case of gray images can be described as black element at zero value, the value was more than increase the severity of lighting element. If allocated for each element Byte one character to represent the light intensity, it can represent the elements of great white Alastdhaep number greater than zero, and represent shades of gray numbers between zero and number the color white.

XI: blending images

Although you can use a single image of the narrow bands and corrected in the processes of interpretation, but that the possibility of blending between images of different scales is more appropriate for an analyst or interpreter in determining the optical data on, depending on the objective of the study to be achieved. Produces color images by blending of three different wavelengths, through the drop across the light sources of colors blue, green and red, where, according to color images in the possibility of distinguishing the best.

Can be divided into three groups:

1. Include the first group on the gray image, which can vary the value of any of the elements of the value of black to the value corresponding to the color white. The input device consists of gray image of the computer, usually from a video camera followed by the continuing signs of an adapter, that converts the signals from the video camera on a matrix of numbers, can be entered into memory, the so-called "memory image." And show the image gray strongly regular lighting and details of analytical acceptable, it usually takes to 256 × 256 element of the image. Consequently, the volume of memory required to store the image of a gray medium is 34 KB [1] K. Byte.

2. The second set of images can be characterization of the computer a set of coordinates and simple geometric shapes, such as straight lines, arcs and circles. Represent, usually, these images of geometric shapes as maps of cities, fields, and forms of buildings, electrical circuits, and others.

3. And described the third set of images as binary: black and white. In spite of this binary image can be represented in the size of small memory, if each element such as a binary number, and therefore can represent every 8 elements in the place of one character. This method is not commonly used, because the computers do not usually allow to deal with the binary number in a manner directly. Therefore, each of these elements, either the stored value in place of character, and either need to build special chambers to enter this binary image.

XII: The device image scanner

This is called the device Scanner, and sets its price, which is described quickly scan a single document, as well as its analytical capacity. Valmash average scans document 8 × 12 inches in less time than a minute, and the analytical capability of up to 255 per inch linear element. This means that the document dimensions of 8 × 12-inch lift on it nearly 4 million items.

When the possibility to collect all 8 elements in the place of one character, the amount of memory required half a million characters. From this it is clear that the requirements for large memory sizes, in the numerical treatment of images, is one of the matters addressed.

XIII: The objectives of digital image processing

And associated methods of digital image processing, the objective of this treatment, and these objectives are in:

1. Image enhancement or modification; to show and demonstrate the relevant information for a better way.

2. Make measurements on the images, and a process of mismatch between the elements of the picture.

3. Known parts of the image.

The treatment here is based on knowledge of initial components of the image, and the objective, usually, the operations of classification of the image data by way of its elements, to know their contents.

XIV: Methods for digital image processing

Take place in digital image processing in one of the following three methods:

1. Handling point, and it handles each element of the image on the unit, without regard to other elements.

2. Handling area, and is used to obtain clearer lines of parts of the image.

3. Handling of the framework, and here are two pictures of consecutive treatment, at least, to show the differences between them. And uses this method of control to know the moving parts. Therefore, and, considering the amount of data for each image, it is clear constant need for means faster processing.

XV: advantages of modern remote sensing images

Characterized by images of modern remote sensing data as a high-quality level to it, has the following characteristics:

1. High degree of differentiation, as has moved from the 57 × 79 m in a generation satellite MSS to 30 × 30 m in a generation satellite TM, and 20 × 20 m or 10 × 10 m in the case of the French satellite "SPOT", then came the boom to provide third-generation high-resolution illustration of the image space This minimizes the floor space, which represents the main point Pixel, hit this area
3 × 3 meters in the satellite Early Birds and 1 × 1 m, and 4 × 4 meters in the satellite Quick Birds.

2. High degree of precision spectral Spectral Resolution, and is intended to narrow the spectral range, or short wavelength, which is which capture electromagnetic waves emitted from terrestrial objects, where the wavelength range varies from 90 to 110 nanometers in the French satellite "SPOT", and this long-range generation the second TM of 70 nm in the moon TM-1 and 250 nm for the moon TM-6. Then came the third generation to take us to the extent and quality of spectral space and spectral and multi-spectral another, and when the use of Alasbactromitr, known as CASI, which depends on the dynamics of the spectral range of channels Spectral Band Range, and dynamics of the degree of differentiation.

3. Multiple channels, or wavelengths, that is by capturing emissions of terrestrial objects, Fbnzerp image data to the Moon "SPOT", we find that it picks up only three waves, while giving first-generation satellites, "Landsat" data on four channels, has increased to seven channels, in the statements of the second-generation TM, The third generation of remote sensing negative, jumps to the number of channels to 545 channel.

4. Third-generation data does not need to make corrections engineering, and do not suffer displacement maps.

5. Possibility of changing the floor area, which represents the main point of the image, by changing the high flying, as well as easily change the number of channels and wavelength, and thus multiple-use areas.

6. Available geographic coordinates of the recent data, thanks to the existence of a cosmic navigation GPS, mobile satellites, and thus make the data of the required quality, which has coordinates of the ground.

XVI: The accuracy of extrapolation of results and ease of

Of course, there are no features or previous capacity combined in one system, where the unique advantage of each system, or a number of previous features. Despite the high spectral resolution, and multiple channels for hardware system Alisbactromitr CASI, but shame on this system the small number of key points, which consists of one picture.

When comparing the data with the modern generation of former generations, in terms of negative space remote sensing data, The evolution of the quality of data, which reflected the concept of digital data processing of satellite data and aerial photographs and methods, which are to scan.

This in turn led to broad areas of remote sensing applications, where he helped to reach the third generation data of multi-channel high, and the rise in the degree of finite difference, to change the basis for digital data processing for remote sensing or the concept.

XVII: A method to read the picture in the different generations

During the first generation, treatment aimed at identifying the nature of things terrestrial, of how different groups through the main points Pixels and similarity, and thus missed the extrapolation of the results of image processing and interpretation to many of accuracy, and increased checks and make sure the ground.

On the contrary, the statements of third-generation, multi-Bmmizzatha, led to the concept of a new generation of digital data processing, a concept that the ability to search for points, or search for a small group of key points, which are equally in the values of digital data, or at least, similar to a very high degree of equality, and therefore provides accurate and easy to extrapolate the results.

XVIII: methods to improve image

Include digital image processing of two basic components, namely: to improve the image, and classification of digital images. And aims to improve image access to view and better visibility to the image content of the information, so the result will be more convenient than the original image for a specific application.

According to the types of images used in the treatment, can be applied to the possibilities of improving and restoring different. Improving the image point is essential for subsequent visual interpretation, as well as classification of digital images improved. And possibilities for improvement of different pictures:

1. Geometric correction

The patch uses geometric Geometric Corrections for images of space, which have geometric distortions as a result of the change in scanning speed, and the change in the high-mounted, carrying remote sensing instruments, and speed.

This is important when debugging a match between the satellite images of the different sensor systems. Mathematically calculated conversion factors, based on which the image is corrected by way of re-arrangement of the elements in the level of the picture, so that the satellite image is a digital image processing is known to the values of the coordinates are correct.

2. Atmospheric correction

In the case of atmospheric correction Atmospheric Corrections, arise on the light rays sent from the dispersal of the sun within the atmosphere, and be more on the air scatter light rays with shorter wavelengths, as in the pictures "Landsat".

These rays disseminated added to the radiation reflected from the surface of the Earth to the spacecraft, causing reduction of the degree of variation within the image space. The aim of this patch to modify the light intensity by deleting the value of X from the values disseminated the picture elements.

3. Enhance contrast Enhancing Contrast Contrast, benefiting increase in contrast between the picture elements, especially those with the shades close, so by deploying degrees lighting of the scene, so covering the entire area of the color black and white to color.

4. Cutting density

and utilized cutting-density Density Slicing to cut the intensity of the color of the image space to a number of slides, each slide so that the limit specified range of the degree of color and gives it to a separate color, this method helps to image processing paint.

5. Nomination

and satellite images contain noise Noise signals lead to distorted image. The aim of the use of filters Filters to reduce this noise, which helps to access the best explanation of the content of the image. In selecting a suitable candidate is taken into account the image content information, and the noise ratio, and the objective of the study, and the degree of discrimination of the terrestrial system to be used.

6. Mosaic image [2]

Through the process of mosaic images are linked to the individual selected to get a picture of one, and after conducting enhance the contrast and the various corrections to be all pictures with a regular appearance and Loney, proximity, and the process by using the link points discrimination ground in areas of overlap between adjacent images.

7. Fit the image channels

aims fit channels to reduce the image, or delete, topographical differences between the images so that similar objects have the same percentage, regardless of changes in lighting. This process is useful for better discrimination of the types of rocks and soil on the satellite images. But it reduces the disadvantages of the contrast ratio between the elements of the picture, as can the use of images of proportionality in order to obtain pictures of colored vehicle.