The aim of Classify Pixel is to reduce the number of image elements and enable users to recognize the specific information of each image clearly. For example, the range of vegetation or the waters distribution in the image processed with Classify Pixel can be easily recognized. In order to make the specific information of images be distinguished easily, Classify Pixel can transfer a color image with multiband to a grayscale image with single band and also reinterpret the raw image with the two colors, black and white. As a result, the display of the image will be simple and easy to identify. Since an image processed with Classify Pixel can only display some unique information, users can use one raw image to produce several Classify Pixel images for different image analysis.

Classify Pixel

Color Map is a corresponding table containing color information, which can be set by users or changed by imported color files. Different images transferred with the same color map can produce image data with identical colors, which can be used with other image processing functions.

Color map transferring is to transfer the corresponding pixel of the color map set by users to transfer an image with single band to a RGB color image or to transfer a color image with different color map to another color image.

Normalized Difference Vegetation Index (NDVI) is another image processing function, utilizing color map transferring. As Normalized Difference Vegetation Index (NDVI) is processed, a gray scale image combining Near-infrared ray and Infrared ray will be produced. Through color map transferring, the gray scale image is transferred to a NDVI corresponding color and the image with vegetation color can be produced.

NDVI color map transferring

Pixel type is as know as bit depth. The quantity of pixel type stands for the amount of color information in each pixel. Pixel Type Converting means converting the bit depth of raw image to the bit depth meeting users need, which is a conversion of pixel status. For example, an 8-bit pixel raster image has two to the eighth power (i.e. 256) kinds of color types, the range of which is from 0 to 255. That is to say, the bigger the bit depth is, the more colors can be displayed; the image display can be much closer to the reality.

The image, each pixel of which is 1-bit pixel type, can be displayed with the two colors, black and white; the 16-color image, each pixel of which is 4-bit pixel type, can be displayed with two to the fourth power (i.e. 16) colors. When the pixel of the image is 8-bit pixel type, each pixel can not only display black and white colors but also 256 different gray colors. On the other hand, each pixel of 256-color image is 8-bit pixel type; each pixel of full color image (16777216-color) is 24-bit pixel type. In SuperGIS Server, the minimum conversion is 8-bit pixel type, and the maximum of conversion is 64-bit pixel type. Moreover, this image processing can be applied with other image processing functions provided by SuperGIS Image Server to enhance the use of the processing function. For example, Pixel Type converting can be applied with color map transferring to add colors to the image.

The main function of Extract Bands is to extract bands from one or more than one raster files with multiband to combine a new image

Extract a single band

The specific band processed with Extract Bands can be used with other image processing functions, such as Convert Pixel Type, to improve the speed of image processing and save the time for data processing and converting.

Grayscale is a basic image processing function to convert a color image with multiband into a grayscale image with single band. Grayscale images are composed of black and white colors of different brightness. Generally speaking, each pixel of grayscale image having 8 bits can display 2 to the eighth power (i.e. 256) kinds of gray colors. Thus, it is called 256 grayscale as well.

Grayscale has different definitions in different fields. In the field of digital image, grayscale image means black and white image, called monochromatic image. However, in the field of computing, grayscale image dose not stand for black and white colors only but includes several various gray colors between black and white, which are different from the pure black color (0) and pure white color. (1). In SuperGIS Image Server, Grayscale offers weight value, grayscale normalization, and grayscale standardization. Users can choose the suitable grayscale methods to meet their own need.

Grayscale

Normalized Difference Vegetation Index is applied to estimate the green vegetation of the target area. The estimation method is to divide the difference between Near-infrared ray and Infrared ray into two, and the estimation results can facilitate users to know the growth status of the green vegetation.

The different spectrum features of Near-infrared ray and Infrared ray are applied in Normalized Difference Vegetation Index; the Near-infrared ray will have higher reflectivity in the area with vegetation, and the Infrared ray will have higher absorptivity (lower reflectivity). That is because green plants with chlorophyll will cause reflection difference. Due to the different features of different bands, users can produce the image contrast with NDVI for applications, such as monitoring the drought in some areas, changes of forests, or predict agriculture development.

Normalized Difference Vegetation Index

Ortho-rectification, a geometric processing technique of remote sensing image processing, can expand an image to some extent to improve the accuracy of the image.

Remote sensing images can be the sources for Ortho-rectification. The images being ortho-recitified can be overlapped or displayed with other GIS data, be the application for GIS development, and be the references for the analysis of changes of some area.

2D Ortho-rectification

Pan-sharpen, an Image Fusion technique, can produce a new image by overlapping a high resolution panchromatic raster image and a low resolution multiband raster image. Thus, the surface and the nearby objects in the image can be distinguished more clearly, and the visualization effect and the color quality of the image can be improved as well.

Pan-sharpen

SuperGIS Image Server provides 3 image merging algorithms to pan-sharpen images, including Brovey transferring, IHS transferring, and Simple Mean transferring, which can be selected by users based on their own need.

Resample can be regarded as a technique for the display of user-end devices, which can be applied to various cases, such as images displaying in different size display devices, and image scaling or resizing in image processing software.

Users can set the type of resampling and change the resolution of the raw image. Different from Ortho-rectification, which needs to take height as one of the 3D resampling sources, Resample can only deal with 2D data resampling.

Since SuperGIS Image Server is a huge image internet server, the Boundary of the image can be set in Resample. To set the boundaries of the image is like to cut the image; with the boundaries set by users, the server does not have to recalculate the whole image but some specific boundary areas.

Resample

SuperGIS Image Server offers 4 resample methods, such as Nearest Neighbor Interpolation, Bilinear Interpolation, Bicubic Interpolation, Majority Interpolation.

Stack Bands is to combine bands of the image to enable users to combine images of different formats or sizes to produce a new image. Users can select specific bands of the image to produce a multiband image or grayscale image.

Multiband image produced by stacked single-band images

Stretching is a technique to enhance images, which enables users to adjust the brightness, contrast, gamma value of the image to expand the features of the image and to enhance the visualization effect for image analysis.

The brightness of raw image is generally from 0 to 255; users can adjust the range to increase the brightness of image.

The lighten image and the histogram

Contrast is the range of the difference between the brightest area and the darkest area. The bigger the difference is, the higher the contrast is, vice versa.

Higher contrast image and the histogram

Gamma value]^^can be adjusted by users need. If users would like to increase the display of the dark part of the image, Gamma value can be adjusted to higher to display brighter image. That is because the bright part of the grayscale is compressed, the brightness of the dark part becomes enhanced. However, this action will influence the details of the bright part of the raw image.

Images of different gamma values

Trend is an image processing method to equalize the color and the brightness of images. For example, as the remote sensing aerial or satellite image is taken, different distances might cause the inconsistent brightness. In the case of image taking vertically, the center of the image will be brighter, and the edge will be darker. Such a problem might cause that these images cannot be applied for analysis or processed. For example, some parts being too bright or too dark may result in imperfect images. Therefore, Trend can modify the problem of brightness and produce the images with more consistent brightness and colors.

SuperGIS Image Server provides two methods, Plane trend and Simple curve trend for adjusting plane surface and curve surface.

Adjusting the brightness of the image with Trend

Visualize elevation provides various visualized effects for elevation data display. With the image processing technique, it will be more convenient to analyze the elevation of the image data.

Digital Elevation Model (DEM) is a surface elevation model, utilizing digital 3D spatial matrix. This model usually takes several points of the same distance as sample points and records the elevation and the coordinates, so this model does not include the surface and building information. In SuperGIS Image Server, Elevation-Coded, Hill Shade, Shaded-Relief, Slope, Aspect, Curvature are provided as visualized image processing methods.

Digital watermark, belonging to the field of Data hiding, aims to prevent digital image data from being copied, distributed, and used without authorization. To embed the related information to the image is a kind of image processing technique. That is to say, watermark technique is a way to protect the intellectual property rights of image data and to prove the legality of data use.

The watermark function in SuperGIS Image Server can embed and extract the invisible watermark; users can embed or extract the watermark of copyright announcement. In order to prevent the raw data from being unable to load huge watermark data, it is recommended that embedded watermark should be black and white; the size of the watermark should be less than one fourth of the original image. Additionally, there is no specific limit for the format of the embedded watermark. On the other hand, when the invisible watermark is being extracted, SuperGIS Image Server can match the original watermark and calculate the accuracy of the extracted watermark to ensure the copyright of the image data.

Extracting the invisible watermark