im_dbr
IM_DBR
Image Database Retrieval GUI
[RANK,TARG,OUTL] = IM_DBR(DBASE,FSETS,CLASSF,COMB)
Input | DBASE | Dataset or datafile with N object images | FSETS | Cell array with maximum 4 feature sets | CLASSF | Cell array with untrained classifiers (Default: KNNC([],1)) | COMB | Combining classifier (Default: MEANC) |
Output | RANK | Index array ranking the N object images | TARG | Index array pointing to user defined target images | OUTL | Index array pointing to user defined outlier images |
Description This command generates a Graphical User Interface (GUI) enabling the user to label a database of images in 'target' and 'outlier' images in an interactive and iterative way. Up to four feature sets can be given and corresponding classifiers that assist the user by predict an object ranking based on classification confidences for the 'target' class.
The GUI shows the top-10 of the ranking and the user should classify them as targets or outliers (original object labels in DBASE are neglected). There are buttons for browsing through the ranked database or through the selected targets and outliers. Classifiers can be trained according to two different strategies using the top right buttons Classify - uses all stored target and outlier objects (shown in the top left windows) for building a training set as well as the hand labeled images in the present screen. Label | - uses just the hand labeled images in the present screen and neglects the stored targets and outliers. This enables a more flexible, but still controlled browsing throug the database.
| Reset | - Resets the entire procedure by deleting all selected targets and outliers.
| Quit | - Deletes the GUI and returns the ranking and selected targets and outliers to the user.
| A | few additional buttons and sliders for controlling the system behavior: | - | Delete and move buttons for the selected targets and outliers | - | Weights for the feature sets. For each feature set a different classifier is computed generating target confidences for all images. This influences the operation of the combiniong classifier. The weights can be changed by a slider for every feature set. By default weights are 1. | - | Two buttons for setting all labels as target ('All target') or outlier ('All outlier'). | - | Labels for the individual images can be changed by a mouse-click in the image or on the image check-box. | - | For all images a target confidence is computed. Depending on the 'all' and 'unlabeled' radio buttons at the bottom the ranking of all images or of the yet unlabeled images are shown. | Note: | It is not an error, but for most classifiers useless or | counterproductive | to label an object as target as well as outlier. |
Example(s)
% This example assumes that the Kimia images are available as datafile
% and that the DipImage image processing package is available.
prwaitbar on
a = kimia_images;
x = im_moments(a,'hu');
x = setname(x,'Hu moments');
y = im_measure(a,a,{'size','perimeter','ccbendingenergy'});
y = setname(y,'Shape features');
[R,T,L] = im_dbr(a,{x,y}); % do your own search
delfigs
figure(1); show(a(R,:)); % show ranking
figure(2); show(a(T,:)); % show targets
figure(3); show(a(L,:)); % show outliers
showfigs
See also
datasets, datafiles, mappings, knnc, meanc, This file has been automatically generated. If badly readable, use the help-command in Matlab. |
|