GC Image Users' Guide

Visualizing Image Data

GC Image has various utilities to view GCxGC image data. Users can:

Resize the Image Viewer window.
View an image at various levels of magnification (i.e., zoom).
Resize an image.
Navigate about an image examining the pixel values both visually and quantitatively (i.e., pan and scroll).
Change the colorization used to display an image.
View data values in a text-format table.
View data with three-dimensional perspective.
View (and save) one-dimensional slices and projections of rows and columns.
Plot various statistical features of the image data.

The following sections describe each of these facilities. Additional visualization tools for GCxGC images with mass spectra (GCxGC-MS) data are described in GCxGC-MS Data. Additional visualization tools for blob data are described in the Peak Detection and Analysis.

Image Viewing and Navigation

The size of GCxGC images typically exceeds the capacity of the computer display monitor. Then, an entire image cannot be displayed at its full resolution. GC Image has two principal tools for viewing an image:

The Image Viewer, described in this section, displays a sub-region of the image at the specified location, scale, and size.
The Navigator, described in the next section, displays the full image at reduced magnification.

When an image is imported in GC Image, the image is displayed in the Viewport of the Image Viewer at 1x magnification with the first data point in the lower-left corner. (See File Input and Output for instructions on opening an image.) At 1x magnification, each data point is displayed with one pixel of the display device.

Figure 1: The Image Viewer with an open image with a sub-region displayed in the Viewport at 4x magnification region of a GCxGC image.

To resize the Image Viewer, position the cursor on a border or corner of the window and click-and-drag with the left mouse-button to a new size. When the Image Viewer is resized, more or less of the image is visible in the Viewport and the magnification (or scale) factor is not changed. The Maximize button on the title bar makes the Image Viewer fill the screen and the Restore down button will restore the previous size after maximization.

As the cursor is moved about the image, the status bar at the bottom of the Image Viewer displays the abscissa (the X coordinate, along the first column dimension, displayed horizontally left-to-right), the ordinate (the Y coordinate, along the second column dimension, displayed vertically bottom-to-top), and the value of the pixel indicated by the cursor. See Figure 1. The axes units can be set to pixel units or time units, via Configure -> Configure Settings on the menu. For time units, the abscissa is in minutes and the ordinate is in seconds.

When an image is opened in GC Image, the Image Viewer is in View cursor mode. To return to View mode from another cursor mode, click the View mode button on the Image Viewer palette. In View cursor mode, it is possible to pan or scroll about the image by click-and-drag with the left mouse-button. For example, to slide the image to the left, depress the left mouse-button (with the cursor in the Viewport), move the mouse left, and release the left mouse-button. The arrow keys also can be used for pan and scroll at increments of four pixels. The page-up and page-down keys move a full screen up or down. (Note that it is necessary to click in the Viewport to activate the listeners for these keyboard actions.) There is also a special navigation tool, the Navigator , described in the next section.

Change the magnification factor with the Scale Viewport control on the Image Viewer tool bar or in the Navigator, described in the next section. To change the magnification in the Scale Viewport control, select one of the values from the pull-down menu or enter the desired value (followed by enter) in the text box. The Scale Viewport menu also has choices to fit the image to the Viewport. The Viewport image is magnified using interpolation. The interpolation method can be set using the pane accessed by Configuration -> Axes. Bi-linear interpolation is the default, but nearest-neighbor interpolation can be useful for making clear the pixel resolution. Figure 1 pictures an open image with a sub-region displayed in the Viewport at 4x magnification.

GC Image also supports more general resizing that changes the aspect ratio. Through the dialog, shown in Figure 2, accessible with the View -> Set Viewport menu option, the user can change the size (including aspect ratio) and position of the viewport on the image used for display operations. Set Viewport operates on the full image buffer that is used in a variety of operations. For example, resizing is useful before exporting an image in external file format. Resizing is different than scaling in the Scale Viewport control, in that the aspect ratio can be changed. Resizing is slightly more time consuming than scaling.

Figure 2: The Set Viewport dialog is accessible from the GC Image menu bar and from a button on the Navigator toolbar.

Resizing and repositioning also can be performed in View cursor mode by click-and-drag of the right mouse-button in the Viewport. Then, the viewport is set to show the delimited rectangular region. Successive resize actions can be performed. The Display Reset button resets the scale to 1 and the size to the base dimensions.

Changes in size and position are stored in a display setting stack. The toolbar has Previous and Next arrow buttons to move back and forward in the display stack.

Navigator

The Image Viewer supports full resolution and magnified views of the image, but only a sub-region of the image may be visible. The Navigator provides a "thumbnail" view of the entire image (typically at reduced resolution). The Navigator also provides an interface to visualize and relocate the ROI that is displayed in the Viewport of the Image Viewer. Figure 3 illustrates the Navigator .

Figure 3: The Navigator with ROI specification of the image in the Viewport.

To open the Navigator, click the Navigator button on the Image Viewer tool bar or select View -> Navigator from the menu. The size of the Navigator can be changed by grabbing and click-and-drag the border or corner of the window. As the size is changed, the thumbnail image in the Navigator is resized to fit the window, changing the magnification and/or aspect ratio of the thumbnail image to show the entire image. The user can indicate whether the Navigator displays the image's original size, the current aspect ratio, or match the size of the Navigator window. These choices are available on the Navigator tool bar. Close the Navigator by dismissing the window.

The graphic rectangular box in the Navigator is the region-of-interest (ROI) that is displayed in the Viewport of the Image Viewer. The ROI can be relocated by using the mouse to grab the box and drag it to another location or via the Position Viewport button on the Navigator tool bar. The ROI can be resized by grabbing an edge or corner of the box and dragging it. The ROI also can be resized by click-and-drag of the right mouse-button in the Navigator. Then, the rectangular region delimited is scaled and resized to fit the Viewport. When the ROI is relocated or resized, the image displayed in the Viewport is changed accordingly. Likewise, the ROI is relocated or resized when the Viewport is moved or the Scale Viewport setting on the Image Viewer tool bar is reset. The Set Viewport button invokes the Set Viewport dialog, described above, which allows precise specification of the size and location of the ROI.

Colorize

GCxGC images have an intensity value (or total intensity value for MS data) for each pixel, with support for full double-precision floating-point representation. A natural way of viewing images with one value per pixel is to use a grayscale, in which values are displayed from black to white in various shades of gray according to their magnitude, e.g., black for the pixels with smallest value and white for the pixels with largest value (or vice versa ). A grayscale provides a clear ordering of values from small to large. Unfortunately, the human eye can distinguish only about 100 distinct gradations on a grayscale, so the large range of values present in GCxGC images cannot be appreciated fully in grayscale.

Colorization allows the human eye to distinguish many more distinct gradations in a GCxGC image. Colorization uses three values to specify a color for each value in the image. Each pixel value is mapped through a color map to a corresponding color, so two pixels with the same values will be displayed with the same color and two pixels with different values may be displayed with different colors. With colorization, the ordering of values is not as straightforward as with a grayscale, but a good color mapping scheme can provide a clear progression. For example, topographic maps commonly use an ordering from small to large that progresses through dark blue, light blue, dark green, light green, light brown, dark brown, and red.

The GC Image Colorize tool allows user control over the color mapping. The Colorize tool supports:

Loading of pre-defined color maps.
Interactive specification and modification of color maps with:
- either curve-fit or linear Color Map functions;
- either Red-Green-Blue (RGB) or Hue-Saturation-Value (HSV) Color Model;
- Value Mapping via a parametric exp-log function or an automated algorithm; and
- Color Wheel rotation.
Saving of user-defined color maps.

Open the Colorize by either clicking the Colorize button on the Image Viewer tool bar or selecting View -> Colorize from the menu. Figure 4 illustrates the Colorize window.

Figure 4: The GC Image Colorize tool.

The Colorize tool has several components:

A sub-region of the image from the ROI is displayed in the upper-left corner of the Colorize tool. The colorization changes can be previewed in this component.

Three Color Map functions, either RGB or HSV, depending on the current color model, are displayed in the middle of the Colorize tool. The Color Map horizontal axes are for the pixel value (after mapping through the Value Map) over the current Value Range. The Color Map vertical axes are the color component value (Red, Green, or Blue for RGB or Hue, Saturation, or Value for HSV). The Color Map functions assign color component values for each pixel value in the Value Range .
The Color Map functions can be changed by grabbing and relocating a knot, indicated with a small square. As the cursor is moved in the graph, the cursor location is displayed on the graph control panel. The function is interpolated between the knots linearly or with a spline curve and the interpolation method can be specified on the graph control panel. Each color map function has buttons to add and remove knots and to reset the function (undoing changes made to the Color Map function).

When the Color Map functions are set as desired, they can be applied to the image with the Apply button at the bottom of the Colorize tool. The OK button applies the current Color Map to the image and dismisses the Colorize tool. The Cancel button dismisses the Colorize tool without applying the Color Map.

A radio button indicates Parametric or Automatic value mapping.
- Parametric value mapping enables a slider to control value mapping. A parameter setting equal to 0.5 is a linear mapping of the range of pixel values to the color scale; a parameter setting less than 0.5 is a logarithmic-type mapping, with color transitions principally in the lower range of values; and a parameter setting greater than 0.5 is an exponential-type mapping, with the color transitions principally in the upper range of values. The Value Map controls include a button to reset the mapping parameter to 0.5 for linear mapping.
- Automatic value mapping uses an analysis of the relationship between intensities and gradient magnitude (in non-excluded regions) to determine the value mapping. This algorithm frequently yields a logarithmic-type mapping for GCxGC data. With automatic value mapping there are no parameters for the range of value mapping (the full range of values is used) nor for the function over the range, so those GUI controls are disabled in Automatic value mapping.
The value mapping is graphed, with pixel value for the abscissa (from the value mapping range minimum to the value mapping range maximum) and the colorization index for the ordinate.

The Value Range for Parametric value mapping sets the range of values that are mapped across the color scale (i.e., values between the Value Range minimum and maximum are color mapped). Values outside the Value Range are set to the Extrema Colors, also specified on the control panel. The labels on the minimum and maximum text boxes give the minimum and maximum values in the image. Note that after entering a desired minimum or maximum value, it is necessary to enter the value by typing the <tab> or <ENTER> (or <RETURN>) key.

The three component color maps combine to form a color bar which can be treated as a Color Wheel, with one end of the color scale wrapped around to other end of the scale. The Color Wheel can be rotated by click-and-drag on the color bar or by specifying the shift (between 0 and 1). The current shift is indicated in the shift textbox label. Note that if the shift is changed by typing in the textbox, it is necessary to enter the value by typing the ENTER or RETURN key on the keyboard. When using the Color Shift, the Extrema Colors (described below) typically also should be set.

The Color Space is specified as either Hue-Saturation-Value (HSV) or Red-Green-Blue (RGB) on the Colorize control panel below the Color Shift.

The color of pixels with values less than the minimum of the Value Range or greater than the maximum of the Value Range can be set as Extrema Colors. The automatic setting is to use the color value of the Value Range minimum value for all values less than the minimum and to use the color value of the Value Range maximum value for all values greater than the maximum. The Extrema Colors controls, located below the Value Map controls, also allow custom settings for these values.

The Configuration tab is one of two tabs located below the Extrema Colors controls at the bottom right side of the Colorize tool. The configuration tab only consists of two buttons, Add and Delete. All operations performed on the color map are applied to the currently selected color map configuration. After the user is done with changes to the color map, click Add and a copy of the color map will be added to the configuration list. Up to four color maps can be stored for a single image. To remove a color map from the list, simply select the configuration and press Delete. After the color map tool is closed, the user can select the image and use the c or C key to toggle through the color maps.

The Import/Export controls allow users to access several color maps supplied with GC Image and to save and load custom color maps. These controls are the second tab located on the bottom right side of the Colorize tool. To import a color map, select the desired map from the list of color maps with a mouse click and then click the Import button. To export a color map, enter the desired name for the color map in the text box and then click the Export button. To delete a color map, select the desired map from the list of color maps with a mouse click and then click the Delete button. The color map named default will be used whenever an image is imported image without a color map.

Text View

The GC Image Text View tool provides a tabular view of the data values, as illustrated in Figure 5. Text View is available as a button on the Image Viewer tool bar and as View -> Text View from the menu. Terminate Text View by dismissing the window.

Figure 6: The GC Image Text View tool.

If a rectangular region is selected using the Graphics tool (as described in Graphics), then only the values in the selected rectangular region are shown in Text View. If one or more blobs are selected, then only the values in the bounding box of the blobs are shown in Text View . Otherwise, the entire image is available, which can entail extensive scrolling.

If blobs have been detected for the image (as described in Peak Detection and Analysis) then Text View attempts to visually highlight the blobs using randomly generated distinct colors for the text for each blob. Otherwise, if blobs have not been detected, the Text View shows all values in black text.

If the "Show blob information" checkbox is checked, Text View also indicates blob IDs and the value percentages for each pixel. Figure 7 illustrates Text View with blob information.

Figure 7: The GC Image Text View tool with blob information.

The array of pixel values displayed in a Text View window can be written to a text file in comma-separated-value (CSV) format by clicking the Save text view button on the Text View tool bar.

Three-Dimensional Perspective View

The GC Image 3D View tool shows a three-dimensional perspective view of the data values presented as a surface, as illustrated in Figure 8, with optional axes, bounding box, and 1D chromatograms projected on the bounding-box backplanes. 3D View is available as a button on the Image Viewer tool bar and as View -> 3D View from the menu. If 3D View is started when the Image Viewer is in Graphics mode and one or more graphics are selected (as described in Graphics), then only the rectangular region bounding the graphics is shown in 3D View. Likewise, if 3D View is started when the Image Viewer is in Blob mode and one or more blobs are selected (as described in Peak Detection and Analysis), then only the rectangular region bounding the blobs is shown in 3D View. Otherwise, the entire image is shown.

Figure 8: The GC Image 3D View tool.

On the toolbar, the 3D View tool has a button for showing and hiding the 3D View control panel and a button for saving an image of the 3D view to a file in BMP format. The 3D View image also can be copied to the system clipboard (from where it can be pasted into other applications) by clicking in the 3D View image and then typing the <F2> key.

The 3D View tool has a checkbox on the status bar for entering and leaving information mode. In information mode, clicking on the 3D surface will provide pixel data and blob metadata (as specified by the Blob Mouse-Over configuration) at the indicated position (which is shown with a vertical bar). If the information mode box is not checked, then the 3D View tool is in navigation mode, allowing for interactive scaling, rotation, and translation.

The 3D View control panel, pictured in Figure 9, provides scroll-bar control of viewing angle and distance, as well as controls for wireframe or continuous surface; for monochromatic or colorized surface (and color form monochromatic colorization); for showing the bounding box, axes, and 1D projections; for value range and mapping function; for vertical scaling, and for resolution scale. Cursor and mouse actions also are supported directly in the 3D View tool: translate the image by click-and-drag with the right mouse-button; rotate the image by click-and-drag with the left mouse-button; and zoom (or scale) the image by click-and-drag with the middle mouse-button. Configure -> Configure Settings also provides access to 3D View configuration settings. The control panel can be dismissed and restarted from the 3D View toolbar. Terminate 3D View by clicking the close button on the control panel or by dismissing the 3D View window.

Figure 9: The 3D View control panel.

Multi-Projections

The GC Image Multi-Projections tool supports graphing of one-dimensional slices and projections of an image. A projection is the one-dimensional sum across a range of rows or a range of columns. A slice is a single row or column (which is the projection with a range of only one row or one column). The Multi-Projections tool also supports multiple slices as described below.

Figure 10 illustrates the Multi-Projections window with graphs of the selected row and column slices.

Figure 10: The GC Image Multi-Projections tool with graphs of the selected row and column slices.

The Multi-Projections tool displays the sub-region of the image from the ROI. Clicking on a location in the sub-image in the Multi-Projections tool specifies the row and column displays:

a one-dimensional graph of the pixel values from the specified row in the panel below the image sub-region and
a one-dimensional graph of the pixel values from the specified column in the panel to the right of the image sub-region.

The location of the sub-image can be moved left or right (or up or down) by click-and-drag with the left mouse-button on the graph panels at the bottom or left of the Multi-Projections tool. The colors of each graph and associated column selector can be set on the Multi-Projections tool control panel. Optionally, the Multi-Projections tool will graph the estimated baseline level for single slices.

Clicking and dragging in the sub-image display defines a range of rows and columns. With a range of rows and columns, there are two options:

Integrated projection: The pixel values from the selected rows in each column are added together, yielding a vertical projection graphed below the sub-image, and the pixel values from the selected columns in each row are added together, yielding a horizontal projection graphed to the right of the sub-image. Figure 11 illustrates the integrated projection.
Multiple slices: The pixel values in several uniformly spaced rows from the selected range of rows are graphed below the sub-image and the pixel values in several uniformly spaced columns from the selected range of columns are graphed to the right of the sub-image. The Multi-Projections tool allows specification of the number slices with the multiple slices option. Figure 12 illustrates multiple slices.

Figure 11: The GC Image Multi-Projections tool with integrated projections from the selected rows and columns.

Figure 12: The GC Image Multi-Projections tool with multiple slices from the selected rows and columns.

The slice, projection, or multi-slices along either the vertical or horizontal dimension can be saved to a text file with buttons at the bottom of the Multi-Projections tool. Other checkboxes indicate whether or not baseline level is graphed and blob outlines are displayed.

The Shape Projection mode in Multi-Projections projects only the pixels in one or more selected graphic Shapes (e.g., polygons or rectangles). The graphic Shapes must be selected before starting the Multi-Projections tool. Figure 13 shows projections for a polygon.

Figure 13: A Shape projection in the Multi-Projections tool.

Visualize Data

GC Image supports visualization of various pixel statistics, as illustrated in Figure 14. If a rectangular region is selected using the Graphics tool (as described in Graphics), then the statistics include only the pixels in the selected rectangular region. Otherwise, the statistics include all pixels in the image. The Visualize Data tool is launched from the Visualize Data button on the Image Viewer tool bar or from View -> Visualize Data menu item. Close Visualize Data by dismissing the window or clicking the close button.

Figure 14: The GC Image Visualize Data tool.

Visualize Data reports several pixel statistics of the reported rectangular region:

the bounding box,
the minimum and maximum pixel values, and
the mean and standard deviation of pixel values.

Visualize Data contains a two-dimensional plot with user-configurable values. The values that can be plotted are:

pixel value,
row and column indices, and
discrete directional derivatives and gradient of value.

Visualize Data also offers a popup with a histogram plot of pixel values, as pictured in Figure 15 .

Figure 15: A histogram of pixel values.

The Visualize Data tool and the histogram popup are closed with buttons or by dismissing the window.

Contents Previous: File Input and Output Next: Graphics