Images in Manifold are usually the familiar sorts of image that almost everyone has viewed or edited with applications such as Microsoft Paint, Microsoft Image Editor, Adobe PhotoShop, Microsoft Photo Editor and similar programs.
Such images are commonly stored as TIF, BMP, GIF, JPG or various other common image formats. Existing images may be imported into a Manifold project using the File - Import - Image command. New images may be created in Manifold and then exported into common image file formats using File - Export.
Creating a New Image
1. Open a project or create a new project with File - New.
2. Choose File - Create - Image to launch the Create Image dialog.
3. Provide a name for the image, the desired size in pixels and the type of image and press OK.
The File - Create - Image dialog takes the projection parameters for the new image from whatever window is active at the time the new image is created. If the project pane is active, the image will be created using the system default projection of Orthographic centered at the 0,0 world latitude/longitude origin. If a map window or other image window is active when the new image is created, the new image will be created using whatever projection is used by that active window. This context-sensitive setting of default projection parameters makes it much easier to create new images using projection parameters that are hassle-free by default.
An image that is imported into Manifold is stored within the Manifold .map project file. An alternative to importing images into Manifold is to link images into the project using the File - Link - Image command. Images may be linked into a Manifold project from a variety of sources outside of the .map file. When images are linked into a Manifold project the .map project file does not contain the actual image but rather contains a pointer to the source of the image, which remains stored outside of the project file.
See the Linked Images topic for additional information.
Images are made up of pixels ("picture elements") arranged from left to right in a series of equal rows. They are often referred to as raster files in older GIS applications. ("Raster" is the old television word for the sweep of an electron gun across a series of rows used to create an image.)
Although images are normally photographic or scanned images that represent visual images of the world around us, images in Manifold can also be abstract data sets that are saved as pixels in a raster arrangement. For example, a multi-spectral satellite sensor might create an "image" of a continent from space where each pixel contains numbers representing temperature, reflectance, height above sea level as well as other information. Pixels in such an image may be colored according to their data values to convey the data content of the raster data set in a visual way.
Let’s begin by explaining images in the context of familiar sorts of photographic images.
Images are fundamentally different from drawings in several ways:
· Images are made up of a sea of pixels and contain no "objects." Any objects we perceive in an image are the sole result of our psychological interpretation of regions of pixels of like color to be distinct things. Zoomed up close the image is a blur of pixels. Zoom far enough into the image above and it is unclear where the "sky" ends and the "monument" begins. In contrast, drawings contain true objects that are defined by precise sets of geometric coordinates for each object.
· The appearance of an image is highly dependent upon the zoom level. At zooms greater than 100%, the individual pixels will become visible as the image becomes ever more like a coarse mosaic. Drawings, on the other hand, are always razor sharp no matter how high the zoom. Objects in drawings will not change their appearance when zooming in or out: if a point is formatted as a small dot the dot will appear to be the same no matter what the zoom. In a drawing an intersection between lines will always be razor sharp whereas in an image what might appear to be a sharp intersection will at high zoom be seen as clumps of pixels.
· Images have no intrinsic coordinate system embedded in the image but rather depend upon an implied coordinate system suggested by the relationship of pixels to each other that is defined by their arrangement in rows. Drawings have an implicit coordinate system implied by the coordinates that define the objects. If an image is to be used within a coordinate system, it must first be referenced to that coordinate system.
· When drawings are re-projected the number of objects stays the same even though their shape may be different. When images are re-projected pixels are always created or deleted through interpolation as the images is projected into a different shape.
Compressed images are a special type of image that is dynamically reconstituted as needed from compressed data. Although they may appear to consist of pixels and may be described as consisting as a certain number of pixels in width and height, they are dynamically created as necessary and do not actually consist of a fixed number of pixels. Compressed images are used for very large images where display must be fast and storage requirements reduced to accommodate the very large size of the image. In exchange for fast speed and reduced storage, compressed images are limited to display only functions and may not be edited.
Georegistering is the process of telling Manifold where a particular image is to be located on Earth. See the Coordinates Tutorial for more information on the concepts behind this process. Georegistration is closely related to the idea of projecting an image.
By default, images in Manifold are taken to be in Orthographic projection located at zero degrees latitude and zero degrees longitude, with each pixel in the image assigned a true geographic location based on the given size of pixels or DPI used when the image was created or imported. It does no harm to assume a geographic location for images that will be used in a purely abstract way (like PhotoShop), and having a built-in assumption about projection and location is very useful when images will be used in a geographic context.
See the Georegistration topic for information on how to georegister an image for use within a map. See the Projections and Images topic for more information on use of projections with images.
Different Types of Images
Because images can involve very large numbers of pixels and each pixel can involve several numbers to describe its color, images can require very large amounts of storage space. Many methods have been invented to save space when storing images. The more popular methods have become so prevalent that most image processing systems, including Manifold, allow working with images that are internally structured using several different standard methods. These appear in the system as several standard types:
· Grayscale Images - These are coded using one number per pixel representing one of 256 different gray tones ranging from black to white.
· Palette Images - These are images coded using one number per pixel, where the number specifies which color in a palette of up to 256 different colors should be displayed for that pixel. The colors in the palette can be True Color RGB colors. Palette images save space at the cost of a reduced total number of colors available for use in the image.
· RGB Images - These images use three numbers for each pixel, allowing possible use of millions of colors within the image at the cost of requiring three times as much space as grayscale or palette images. They are often called True Color RGB images in Microsoft applications.
· RGBa Images - These images are RGB images with a fourth number added for each pixel that specifies the transparency of that pixel in the range 0 to 255. RGBa images are used when combining multiple images in maps for elaborate graphics composition or creation of special visual effects in maps.
· Compressed Images - Compressed images use sophisticated wavelet compression technology to not only compress the amount of data an image requires but also to reconstitute the image on the fly on demand. At any given zoom level the desired view of the image is reconstituted from the compressed data store. Compressed images in general may be viewed but not altered or otherwise manipulated.
In addition Manifold can import data from multi-spectral raster data images that contain many channels. When importing from formats that support many channels Manifold will import each channel into one of the above image types (most normally, as a grayscale image). Operations on multi-spectral raster data sets can then proceed by choosing those images/channels to use. For example, three of the imported images can be combined into a single RGB image to create a "false color" image that uses three images/channels as R, G and B channels in a single image.
For many operations the various types of images will seem equivalent, except that compressed images cannot be edited or manipulated. We can select a rectangular region of pixels in an non-compressed image using Select Box, for example, without it mattering what type of image is involved. Some operations will only work for certain types of images. For example, the Hue / Saturation command only works with RGB or RGBa images and so this command will be disabled whenever the focus is on a grayscale image.
We can always convert a grayscale or palette image to RGB in order to use a particular command and then convert it back. To convert an image from one type to another, use the Image - Convert To dialog
Transparency in Images
Drawings consist of empty space in any region that is not occupied with a point, line or area. In contrast, images have no empty space. Every part of an image is filled with pixels. To allow "see through" regions in images, Manifold images may have invisible pixels through which any items in lower layers may be seen. Invisible pixels are simply placeholders that do not appear. It is if there are no pixels in that part of the image. Invisible pixels are often used to give the appearance that images are not rectangular but consist of some irregular, non-rectangular shape. However, every image is rectangular because every image consists of a series of equal-width rows of pixels.
To make part of an image transparent, we select the desired region and then Delete those pixels. Invisible pixels may be used with any of the four types of images.
Any image layer can be made partially transparent by changing its opacity from 0% to 100% in steps of 1 percent using the controls in the Layers pane . This is a great way to create spectacular effects. Transparent layers work with all types of images.
In addition, RGBa images can have each individual pixel assigned a percent pixel transparency value. This is normally accomplished through partial erasers and other tools. When combined with Manifold image editing tools this effect can be used to compose amazing images by combining many layers of other images. See the Layers topic for an example.
To summarize, there are four types of transparency that may be found in a map:
· Drawings are transparent in empty space not occupied by an area, line or point.
· Any image may be made fully transparent in regions of invisible pixels.
· Each pixel in an RGBa image may be partially transparent using RGBa pixel transparency.
· Any image layer may be made partially transparency using layer transparency.
Image Windows vs. Map Windows
An image may always be popped open in an image window to show the image by itself using the native projection used for that image. This is a handy way to see the image in its native state if there is any confusion about how it appears in a map.
Map windows are the normal user interface for working with images mainly because of the layer capabilities of maps. Maps can include many image layers and thus allow the use of many simultaneous images. In graphics composition we will often work with many image layers at the same time to compose an image by stacking elements.
The illustration above shows four image layers above several drawing layers in a map. The Manifold logo and text are both in separate layers and their shadows underneath are drop shadows in RGBa image layers created using Gaussian Blur . Both drop shadow layers have had opacity decreased so their shadows are not so obvious.
Using the Layers Pane with Images
The Layers pane is used to control the appearance of images within image windows. The layers pane includes checkboxes for two system "layers" - a background color layer and a border layer that shows an enclosing box about the height and width of the image.
By default, images are shown using the checkerboard background Manifold uses to provide a backdrop for any transparent regions. The layers pane is shown to the right of the image window. Illustrated is the standard bronze image with regions other than the monument selected and then deleted into Invisible Pixels .
Checking the Border box will draw a one-pixel border around the height and width of the image. This is a good way to see the actual extent of an image that contains regions of invisible pixels.
Checking the Background box in the layers pane will replace the checkerboard background with whatever is the default background color. This is a good way to see the actual extent of an image that contains regions of pixels that are the same color as the background (usually white).
Note that only maps can have true "layers" in Manifold in the sense that they can layer more than one component within the same map window. The border and background "layers" in the Layers pane for images are not true layers even though they appear in the Layers pane in the same manner as do layers in maps. These are simply system controls that take advantage of the Layers pane as a conceptually convenient user interface.
Checkboxes show above other than Border and Background are explained in the Images and Channels topic.
Layouts and the Layers Pane
If an image has any Layouts created they will appear as "layers" in the layers pane for that image. Checking the box for one of these print layout layers will cause a layout rectangle to appear in the image that shows the region covered by the layout.
Right clicking onto the hatched border of one of the layout rectangles in the image will cause a context menu to appear with controls based on that layout rectangle. For example, we can Zoom to a given layout rectangle, Print it or change its Properties. If a layout is empty (for example, if the layout scope is set to selection and no pixels are selected in the parent image) zooming to the layout will do nothing.
Use Tools - Options - Colors - Layout Rectangle to change the color in which layout rectangles are shown. The default color is black.
Transparency and opacity are two terms that mean the same concept viewed from different directions. When something is completely opaque it is not at all transparent. When something is perfectly transparent it may be said to have zero percent opacity.
Which word is used depends on the discussion. When imagining layers stacked up above each other like transparent sheets it is conceptually clearer to use the word transparency. When discussing a specific percentage of light transmission to be applied via a slider bar in a dialog most applications use the word opacity.
The convention in the graphics arts editing software industry is to adjust layer transparency with controls that set a number from 0% to 100% opacity, so that an image with 100% will be fully opaque and not allow any view of an image underneath it. Manifold follows this convention. This convention persists in the graphics arts industry even though the technical implementation of transparency effects is done using an alpha channel within RGBa images where the higher the value of the alpha channel (from 0 to 255) the higher the transparency.
One therefore encounters the slight conceptual dissonance of increasing opacity with higher numbers (up to 100%) in dialogs and other user interfaces while the internal data sets use numbers (alpha channel values) in which opacity decreases with higher numbers. Since we rarely set alpha values by hand this is not so bad. Alpha values are normally set using various tools, such as erasers, or masks. In the case of masks, the darker the mask region the lower the alpha value is and thus the higher the opacity. From a casual conceptual view this is very acceptable because it leads to an effect where black regions of masks cause full opacity and white regions of masks cause full transparency. Since we are used to thinking of "white space" as being transparent this works well as a natural mnemonic for the effects of masks.
Images can be Huge
Images are often saved on disk in compressed file formats such as .tif or .jpg or even within a project saved in Manifold's compressed .map format. However, once the image is loaded into Manifold it must be uncompressed so that every pixel is available for viewing, selection and possible manipulation or alteration as a result of editing or georegistration or other processes. A 50MB, compressed .tif file can easily be expanded into a 150MB uncompressed image within Manifold when it is uncompressed from the storage format.
One way to get an idea of the active size of images is to consider their size: RGB images require 4 bytes for each pixel. A 6000 x 6000 pixel RGB image therefore requires 144 megabytes. When working with four such images at once we will have almost 600 megabytes of image space even before any temporary files are considered.
One strategy to manage very large images is to use compressed images , a special type of image that always remains compressed and is dynamically reconstituted as necessary. Compressed images require much less space either in the Manifold project or when saved on disk, and they allow very fast redisplay. However, the cost of using compressed images is that they can only be displayed - they cannot be edited or otherwise manipulated.
Unless the images are compressed images, to work with large images you must have large amounts of RAM and large amounts of free space on disk. You should also use a reliable operating system such as Windows XP or Windows 2000 that can adequately handle large amounts of RAM. Older Windows editions such as Windows '98 have so many memory-management bugs that they cannot reliably be used for tasks that require large amounts of RAM. Since RAM is very inexpensive, load your computer with enough RAM so that the entire project can fit in RAM memory. Ideally, load a full two gigabytes or more.
For Manifold to work your machine must have adequate RAM, an operating system that can handle the necessary RAM, or adequate disk space. If any of these factors are not present your system will report various error messages when working with images or other large tasks.
Other Image Topics
Projections and Imported Components - Discussion of projections issues and formats when importing images plus simple explanation of channels.
Palettes - How palette images work.
RGBa Pixel Transparency - RGBa explained, with examples.
Projections and Images - More advanced discussion of projection issues with images. Read after reading the Projections and subsequent projection and coordinates topics.
Compressed Images - Compressed images are used for display of very large images.
Linked Images - Linked images are dynamically created from queries or tables.
About the Sample Images provides notes on the bronze monument and other sample images.