Image View

For image viewing and to perform measurements, the Barista tool bar contains different buttons to activate operation modes (Digitize, ZoomIn, ZoomOut, Pan as well as four different monoplotting modes). Zooming in an image view is also possible with the mouse wheel. Pressing the middle mouse button will temporarily switch into pan mode; releasing the middle mouse button switches to the prior selected operation mode.

Digitizing, Measure and Image Matching

To select between Digitize, Measure and Matching mode use the toolbar.

Switching from Digitize, Measure and Matching mode via toolbar

Digitizing and Image coordinate measurement

The Digitize mode is for image coordinate measurement and is activated from the Toolbar via this toggle button:

In digitize mode the measured image points are automatically labeled. This point label can be set by hitting Key L on the keyboard, which starts a Label change dialog. Points measured after setting the label (by pressing Ok or just pressing Enter) have the same prefix to which an incrementing number is added. In digitizing mode the change label dialog offers the possibility to set the current point label for all images, which can be quite helpful for stereo or multiple image measurement. Another way of selecting the current point label is given by the Object Selector.

If the active image view has an asssigned TFW node, the image coordinates of the measured point are also used to determine the planimetric position of a georeferenced point, which will be added to the 3D point node of georeferenced points.

Change Label dialog, in digitize mode, hit Key L

Pressing Backspace and Delete will delete the last measured point. The label of the deleted point will be kept and used for next measured point.

Measurements from georeferenced images

The Measure mode is for the measurement of coordinate differences and is activated from the Toolbar via this toggle button:

The measure mode is to measure two points from one or two georeferenced images and to display the differences in their planimetric position. After two points are measure the coordinates, the differences and the planimetric distance is displayed in a protocol viewer. The measured image points and georeferenced points are saved to the images and the georeferenced point node. Pressing Backspace and Delete will delete the last measured point. The label of the deleted point will be kept and used for next measured point.

Matching of image points

The Match mode is for the measurement of matching image points and is activated from the Toolbar via this toggle button:

The matching mode is for matching of points in all images of the project that have a current sensor model. To set the project's matching parameters use Menu Settings or use the short cut Key O to start the Matching parameter dialog.

Matching parameter dialog

Matching is based on multiphoto least squares matching using object space constraints. That is, the positions of the reference point in the search images are determined so that the homologous image rays intersect in object space, and the 3D coordinates of the object point are estimated in the matching process. (This is the reason why only images having a current sensor model take part in the matching process: otherwise, the image ray could not be determined). The following parameters can be set:

• Images: Here the user can restrict the set of search images that are to take part in the matching process by the band configuration. If same band count only is selected, only images having a current sensor model and having the same number of bands as the reference image (the image in which a point was digitized by the user) are used for matching. Thus, matching could be restricted to panchromatic or multispectral images only. If all images is selected, no restrictions are imposed by the number of image bands.

• Image Depth: Here the user can restrict the set of search images that are to take part in the matching process by the image depth. If identical bpp only is selected, only images having a current sensor model and having the same image depth as the reference image (the image in which a point was digitized by the user) are used for matching. Thus, matching could be restricted to 8 bit or 16 bit images only. If use all bpp is selected, no restrictions are imposed by the number of image bands.

• Transformation and display: If the TFW checkbox is ticked, geocoded orthorectified images could be considered in the matching process, too (not yet implemented). If the show match window checkbox is ticked, a box showing the extents of the reference matching patch centered at the current mouse position will be superimposed to the image.

• Patch size: This is the size of the reference matching patch in [pixel].

• Min Rho: Minimum correlation coefficient of grey levels that is accepted. After least squares matching, the reference patch is transformed into each search image, and for each search image the correlation coefficient of the grey levels is computed. Only for images with a correlation coefficient larger than this threshold, the matching result will be accepted.

• Number of Iterations: Maximum number of iterations in the least squares matching process.

• Stochastic model: The a priori standard deviations of the individual groups of observations used in the least squares matching process:

  • Sigma grey value (8 bit): the a priori standard deviation of a grey level in an 8 bit image, in [grey levels].
  • Sigma grey value (16 bit): the a priori standard deviation of a grey level in a 16 bit image, in [grey levels].
  • Tolerance to epipolar line: the a priori standard deviation of a geometrical image observation, i.e., of an observed image coordinate, in [pixel]. If this standard deviation is small, the matching results will be forced to be on the epipolar line. If it is larger, the geometrical constraints will be less strictly imposed.
  • Height range: Min Z and Max Z are the minimum and maximum height of any point in the scene. These values are used to restrict the search space for matching.

If in matching mode it is possible to display the current patch size used for matching. This can be switch on and of via the Matching parameter dialog or by using the short cut Key M.

After clicking an image point in one image, a matching process is started to find the matching image coordinates in all images of the project. If the matching was successful in at least one other image, the found image coordinate is added to the assigned 2D point node of the image. Automatically a 3D point is calculated and added to a 3D point node with the matched points. The coordinates of the matched point are displayed in the Matching Dialog.

Matching of points in two images, with display of matched 3D position

Pressing Backspace and Delete will delete the last measured image point, all corresponding matches found in the other images as well as it removes the 3D point from the matched 3D point list. The label of the deleted point will be kept and used for next measured point.

Selection and processing of a region of interest (ROI)

The image view also offers the possibility of processing a region of interest (ROI).

Image view, ROI selection

Therefore an ROI can be selected by pressing the left mouse button and dragging. A blue dotted rectangle shows the current selected area. With release the left mouse button the selected area is selected as ROI marked with a blue rectangle. Right mouse click opens an image view context menu. To this ROI, different operations can be applied:

Crop ROI

The crop dialog enables cropping of imagery and DEM data sets within Barista. Cropping of images and DEMs is possible from image views and DEM views after the selection of a region of interest (ROI). After the ROI selection the popup menu contains the function Crop Image/Crop DEM, which opens a cropping dialog. Considering images cropping of assigned node like image coordinate measurements and the different types of sensor model can be applied.

The crop dialog is generated for the Barista View where the ROI was selected. The crop setting becomes available as soon as the assigned node has the according sub nodes. The image below shows the crop dialog with the full set of possible selection options, ticking the check box will crop the assigned node and adds it to the crop result.

Crop of Image/DEM Dialog

The values of the selected ROI are shown in the dialog, displaying the upper left, lower right corners as well as width and height. If the ROI was selected from a view with an TFW assigned to it, the display can be switched to metric values (which is probably useful for cropping orthoimages). These values can be modified within the limitations of the view by the spin boxes or by typing the demanded values. The display of the ROI will be updated accordingly. The pixel display will only show full pixel coordinates, whereas the metric display shows some decimals. The image contents will be cropped to the full pixel though.

Ticking Retrieve measured Image Points will create a Image Point Node assigned to the cropped image node. The points within the area of the ROI are corrected by the crop shift, the point label and e.g. coordinate sigmas remain the same.

Ticking Crop RPC parameters will create a RPC Node assigned to the cropped image node. The parameters of the RPCs are modified due to the crop shift, effecting only the 2D offset of the original RPCs. Other values and sigmas remain unchanged.

Ticking Crop TFW parameters will create a TFW Node assigned to the cropped image node. The parameters of the TFW are modified due to the crop shift, effecting only the 2D offset of the original TFWs. Other values remain unchanged.

Ticking Crop PushBroomScanner parameters will create a PushBroomScanner Node assigned to the cropped image node. The parameters of the PushBroomScanner are modified due to the crop shift, effecting only the camera specific to the resulting image. The shared components of the PushBroomScanner data remain unchanged.

Ticking Crop Affine parameters will create a Affine Node assigned to the cropped image node. The parameters of the Affine set are modified due to the crop shift, effecting only the 2D offset of the original Affine set. Other values and sigmas remain unchanged.

Ellipse fitting

Ellipse Fitting will start an 2D ellipse best fit from the points in the selected area.

Image view, Ellipse fitting from points of selected area

If the ellipse fitting was successful, the user is asked to insert a label for this ellipse. If the assigned image had no 2D ellipse node before it will then be created. Otherwise the ellipse will be added to the existing XYEllipses node.

Image view, display of Ellipse fitting result

In addition the ellipse center is automatically added to the image coordinate node of measured points of the assigned image. If the corresponding 2D ellipse is measured in at least one other image, its image coordinate will participate in bundle adjustments and forward intersections.

Delete Points

Delete Points will delete all image points in the selected area (Caution: no undo or further warning!).

Image view, Deleting of points of selected area

Feature Extraction applied to ROI

Extract Features applies an automatic feature extraction to the ROI after setting the parameter in a feature extraction dialog. Successful feature extraction will add two nodes to the assigned image node; One node will contain the XY Contours, which are basically the extracted edge pixels and one node with 2D Lines, which result from a thin-out process of the edge pixels.

Image view, Feature extraction extraction

To Feature Extraction

Orthogeneration

The function Generate Orthoimage from ROI becomes available if the project contains at least one DEM node and one DEM is set as active. Generate Orthoimage from ROI will start an Orthoimage generation dialog. Orthoimage generation requires a sensor model assigned to the image as well as a DEM. Generate Orthoimage from ROI will automatically determine the extents of the resulting orthoimage from the selected ROI.

To Orthogeneration

Setting Display - Channels, Resampling and Enhancement

Set Display Options will open a Display Setting dialog. The dialog gives access to different types of image view settings such as band order selection, resampling method selection and image enhancement function selection. If an ROI is selected all image enhancement functions are applied locally. For that reason the local histogram will be computed when opening the dialog. See Image View Display Settings for details.

To ImageNode

Back to the Main Window