Start Velocity image program by entering the following at the %
prompt:
/usr/velocity/bin/imageStart Velocity display program by entering the following at the % prompt:
/usr/velocity/bin/display
When you start Image, three windows will appear: the control panel, image display window, and filmstrip window. This is image's user interface with several tools.
Menu Bar - The menu buttons across the top of the control panel are used to select the major functions of File I/O, View modes, and program Preferences.Tool Bar - Just below the menu bar are shortcut buttons for file Open, Save and Close, a
Stop button for interrupting operations, and an Exit button.Action Groups - Toggle buttons labeled Gray, Filter, Logic and Surfer represent the major action groups of image processing and 3D reconstruction operations. Choosing one of these items will display the particular control panel, or toolbox, for that group of actions. When the program starts it displays the Logic Operations toolbox.
Action Options - A toolbox for an action group.
Log - A text based log of operations performed during a session.
As image data are loaded, each is displayed in the image display window. After
loading is complete, the first image in the list will be shown. The image display window can be resized to scale the view of the image up or down. However, for fast image display, it is recommended that you keep the size of the image display window fixed at the dimensions of the image data you will normally be viewing.Instead, you can use the zoom and pan controls in the image display window to
magnify and reposition the image.If the image display window has been resized, you can revert it to its original size by clicking on the Window Reset button in the top toolbar.
Thumbnail representations of each image in the data set are displayed in the filmstrip window. Initially, four to five thumbnails are shown, beginning with the first image in the set. To view additional thumbnails you can use the scrollbar at the bottom of the window, or you can resize the window to view more or less of the set.
Clicking on the Logic toggle button in the control panel's top toolbar selects the Logic Operations toolbox.
Logic operations constitute some of the more powerful features of Image. Several of them, such as Flood Fill and the interactive drawing tools, are used to create region-of-interest (ROI) masks for an image. Others are used to operate directly on mask data themselves, such as the Dilate and Erode operations.
ROI masks are essentially 1-bit per pixel images of a size given by the x, y dimensions of the parent image. Each bit in a mask is used to record a logical 0 or 1 value. There are eight such masks, designated Mask One through Mask Eight, stored in one mask file for each image. They are available for storing up to eight independent logical results per image. For example, the results of thresholding an image could be stored in ROI Mask One, while the results of another operation on the same image, such as Freehand Paint, could be saved in ROI Mask Two.
Image automatically creates a mask file for each image in your data set. This file will be named by adding the extension .msk to the name of the parent image file. The Image program automatically loads each image's mask file as needed.
Mask Operations
Mask data are initially created as overlay graphics in the image display window using a function such as Flood Fill. They are then read into and out of the
overlay from the ROI mask bitplanes using the Mask Ops functions Read Mask, Set Mask, and Set All. The overlay and mask bitplanes are cleared using the Erase Overlay, Clear Mask, and Clear All functions.
Selecting the Gray button in the control panel toolbar displays the toolbox for the Gray Level Point Operations Buttons for selecting the gray level functions are found in the toolbar on the left of the panel, and those for selecting ROI masks on the right.
Across the bottom are buttons for performing an operation (Perform), restoring (undoing) an operation (Restore), and a button for saving results (Save Image)
In the center of the panel is a plot of the gray level histogram of the current image. The ordinate is in units of numbers of pixels (scaled to the maximum gray level), and the abscissa is gray level in the range 0 to 255 for 8-bit data, or 0 to 65536 for 16-bit data. Histogram statistics are displayed across the top of the histogram plot, including mean gray level (Mean), standard deviation of gray levels (Std), the minimum (Min), and maximum (Max) gray levels in the image and the numbers of pixels (NPixels) used to calculate the histogram. If an ROI Mask is being used, the histogram and statistics are computed for pixels within the ROI. In this case, the numbers of pixels, NPixels, gives the
area of the ROI.If an ROI mask other than None is specified, all of the gray level functions confine calculations to just those pixels that have their corresponding mask bit set ON. Otherwise, the operation is performed on the entire image.
Linear Filters - DefinitionsThe filtering operations provided in Image include both linear and nonlinear functions. Among the former are image smoothing, sharpening, and edge enhancing filters. Non-linear operations include median filtering (for removing speckle noise), and local gradient detectors such as the Kirsch, Sigma, Roberts and Sobel operators.
A linear filtering operation is one in which the output image is related to the input by a discrete convolution of the input with a filter or kernel function.
The Norm is a normalization factor that constrains the result to the range of
gray levels of the input image, Bias (a value between 0 and 1) represents an offset to be added to the result, and the summation extends over the size of the kernel. The value of Norm is calculated internally by the convolution routine in the Image program. If the sum of the coefficients in h(i, j) is nonzero, Norm is equal to the reciprocal of the sum; otherwise, Norm is set to one.Filter Controls
Selecting the Filter toggle button in the control panel toolbar displays the Filter Operations toolbox. Clicking on one of the buttons in the left-hand toolbar selects the operation to be applied to the current image. Other controls include buttons for mask selection (ROI Mask), performing the operation (Perform), restoring the original image (Restore), and saving the results of the operation (Save Image).
Filtering can be restricted to an ROI by first using the Logic functions to create and save a mask in one of the bitplanes One through Eight, and then toggling the mask ON in the ROI Mask panel.
All results of filtering are stored in a temporary image buffer such that successive operations can be applied to previous results. You can restore the original image at any time by clicking on the Restore button or permanently save results in memory by choosing Save Image.
Three-dimensional models in Velocity 2 Pro are triangulated surface meshes that represent either iso-gray level values, 3D boundaries defined by masks, or combinations of the two.
Surfaces are closed, topologically connected, 2-manifolds, i.e., they have no holes and the edges of surface triangles are shared by exactly two triangles. These properties allow for a number of operations on surfaces by the software, including surface smoothing, decimation, identification of outer and inner surfaces of 3D objects, and measurements including distances, angles, areas, and volumes. The algorithms used in Surfer to reconstruct surfaces can be summarized as follows:
Images are processed four at a time. The center two images are scanned left to right and top to bottom by a sampling cube whose corners are formed by four pixels in one image and the corresponding four pixels in the next. At each position of the cube the image and/or mask data are tested to determine whether or not the cube is straddling one or more surfaces.
If a gray level threshold alone defines an iso-surface, the gray levels at the corners of the cube are examined to determine if they are above or below threshold. If all eight are either below or above threshold the cube is outside or inside the object, respectively, and no surface triangles are generated. If the cube is found to be straddling a surface, the triangular facets that describe the surface, are generated. Locations of triangle vertices along each sampling cube edge are computed using linear interpolation of gray levels at the corners of the cube, meaning that surfaces can be located to sub-pixel dimensions within the gray level resolution of the image data.
If an iso-surface is defined using both gray level threshold and mask definitions, the same procedure is followed with the added condition that any sampling cube corner pixel that is above threshold also be contained in the ROI. Otherwise, that corner pixel is outside the object, having an effective gray level
of zero.When masks alone define iso-surfaces, the mask data themselves are scanned for surfaces in lieu of the image data. Because gray levels are not used, surface triangle vertices are located to the nearest point of mask/sampling cube edge intersection, and the pixel dimensions give the maximum resolution of the reconstruction.
Display also includes functions for reducing the numbers of triangles in a 3D model, and for saving model data in stereolithography (.stl) format suitable for input to most rapid prototyping systems.
When you start display, two windows will appear: the control panel and the
image display window. This is image's user interface with several tools.
The Display program opens two windows, a main control panel window and
a 3D graphics display window.
Along the top of the control panel are buttons for the File I/O pull down menu, for opening, saving, and closing 3D model files, for exiting the program and for accessing context sensitive help. Additional buttons and text fields are provided for setting/displaying the current directory path, for selecting models to be viewed and for choosing the current models surface material.
Further down in the control panel is a toolbar containing the four buttons: Render, Measure, Edit, and Animate that select one of the four main sets of tools in the Display program. When the program is first started, the Render toolbox is active.
You can readily translate, rotate, or scale the view in the display window by using the left mouse button. Place the cursor in the display window and then press and hold the left mouse button while moving the mouse. The default transformation is to rotate the models, with mouse movement simulating a trackball. Other transformations are selected from the buttons in the top-left toolbar of the display window.
While the mouse is moved, models are drawn at a reduced resolution as determined by buttons in the top-right toolbar - either a wireframe view, a low resolution volume element (voxel) view, or a medium resolution voxel view, as described below.
Clicking on the Render button in the control panel, selects the rendering functions toolbox which is used to control parameters such as lights, background color, positioning of the scene (centering, translate, rotate and scale), anaglyph display, and stereo viewing mode.
The rendering functions include controls for selecting lights and background colors, for capturing the screen to a disk file, and for setting the orientation and position of currently selected models. In addition, you can apply surface smoothing to models, save models in their transformed positions and display them in two forms of stereo: red/cyan anaglyph form or full stereo mode using CrystalEyes2(r) stereoscopic viewing glasses.
Clicking on the Measure toggle button selects the measurement functions toolbox. Before you can make measurements on a model, Display must analyze your model data for surface relationships. If a dialog is displayed to this effect, click on the OK button to proceed.
The measurement toolbox should now be displayed in the control panel.
Buttons are provided for selecting the different measurements of distance, angle, area and volume. You can make measurements on single surfaces or 3D objects, the latter being defined by a single outer surface that contains one or more internal surfaces. To assist in visualizing internal structures controls are provided for selecting them by nesting level and by using transparency to see through outer layers. Clip planes can also be used to expose internal structures.
Clicking on the Delete toggle button selects the toolbox for deleting surfaces and 3D objects. Single surfaces and 3D objects can be deleted by picking them using the mouse. Any number can be deleted within a specified range of volumes by setting interactive controls for minimum and maximum volume.
By using the View Mode controls, surfaces and 3D objects at specific levels of nesting of outer and inner surfaces can be selected for deletion. This feature is especially useful for removing inclusions and holes in objects when only primary outer surfaces are needed in the reconstruction.
The histogram displays a plot of the frequency of occurrence of surfaces or 3D objects in the current model at the selected view level. Values along the abscissa of the histogram vary from the smallest volume to the maximum volume of objects at the current view level. The abscissa is divided into 25 equal-width "bins" for tabulation and plotting purposes. Minimum and maximum volumes are displayed as the Min and Max values at the top of the plot, along with the mean (Mean) and standard deviation (Std) of volume. Numbers of surfaces or 3D objects at the given view level are displayed in the "# at level" text window to the right of the histogram.
The Min Limit and Max Limit sliders are used to select objects for deletion according to the state of the toggle buttons Delete <, < Delete >, and > Delete. Values of the sliders, in units of volume, are displayed in the text fields to the right. Except for the Pick Del function, any surface or 3D object whose volume falls within the selected limits will be deleted when the Perform button is pressed. If Pick Del is selected, individual objects are deleted by clicking on them using the middle mouse button.
These tools are used to reduce the total number of surface triangles in a model by merging small adjacent triangles in relatively flat portions of the surface into larger triangles. Merging is control led by specifying a flatness factor, which is the deviation in surface normal vector amongst adjacent triangles. Where the local flatness is less than the specified amount, the program merges triangles into larger ones. As a consequence, file sizes, rendering times, and prototype building times are all reduced.
To run the triangle reduction function, click on the Reduce button in the toolbar. Sliders along the top control the flatness factor and the terminate count, which specifies the point at which the function should stop.
Flatness is defined in units of angular degrees. If the value is 0 (i.e., no deviation in surface normal vector), the reduction will occur only in areas of the surface that are perfectly flat. Larger values will cause merging of small triangles where the deviation in local surface normal vector is less than the specified amount. Flatness factors in the range 0 to 70 degrees can be specified.
Use the simple animator in Display to make movies of your model(s) spinning around the Y-axis. Set the specified rotation angle increment and number of frames. At each angle the screen image is captured and saved to disk. The frames can then be loaded into your favorite movie editor to create the animation.Pressing the Stop button stops any current animation preview or recording and resets the viewpoint. Clicking on Pause suspends the current preview or recording. To resume, click on the Preview button. The Preview button steps the animation through the rotation sequence using the current rotation increment. Pressing the Record button begins frame-by-frame screen capture of the rendered model(s) as the scene is rotated. Each frame is saved to disk in the specified image file format.
When specifying the total number of frames to capture, the number should be one less than that required to complete one full revolution, to avoid duplicating the first and last frames.
The Help menu provides context help and program information functions. Context sensitive help is available by clicking on the question mark button. In this case, place the ? cursor on any button, label or window and click once more to access a brief help pop-up.
To quit the current image or display session and return to the operating
system, select File -> Exit in the control panel window.
| This information is available in alternative formats upon request by individuals with disabilities. Please send email to alt-format@msi.umn.edu or call 612-624-0528. | ||||||||||||
|
HOME
| QUESTIONS | FEEDBACK
|
||||||||||||
|
||||||||||||
