Background Removal Panel
The background removal panel is one of the left interchangeable panels, it provides three different algorithms for background removal. The image has to be as white objects on dark background (use Invert image command in the Image menu). These tools were designed to work with 2D time lapse movies obtained with light microscopy.
Back to Index --> User Guide --> Panels
Contents
General notes on usage
- Start filtering by pressing the Remove button.
- Check the Test checkbox (9.) for the test run and press the Remove button.
- Select the All checkbox (8.) to subtract background from all images in the dataset.
Morphological opening and background removal
Select the Morph opening checkbox (1.) to select this method.
Original image:
How it works:
1. Generate background image by opening the image with imopen function and strel size defined in the Strel size editbox (4.). This would remove all objects smaller than the strel size from the background image. Use high values in the Strel size editbox.
Background to be subtracted
2. If the Smoothing checkbox is ON (3.), the program filters the background image with gaussian filter with Size and Sigma parameters (6.).
Smoothed background
3. Subtract the background image from the current image.
Result of background subtraction
4. Show profiles for the specified line, if 7. is checked.
Background approximation based on local minima
This method uses extrema and extrema2 functions written by Carlos Adrian Vargas Aguilera from Universidad de Guadalajara, Mexico. The method finds local minima and then generates background image using the amplitudes at the local minima. Select the Minima checkbox (2.) to select this method.
How it works:
1. Prefilter the image with Circular averaging filter (fspecial function with disk type, using size from the Strel size editbox (4.). Increase of the strel size parameter will reduce number of local minima.
Prefiltered image
2. Find local minima with extrema2 function.
3. Remove all local minima that are higher than the Max minima parameter (5.).
4. Reconstruct the background image
5. Smoothen the backgound image if the Smoothing checkbox is selected
6. Generate averaged background image for the whole stack.
7. Substract averaged background from each image in the stack. The local minina that were used for reconstruction are marked with crosses.
Estimation of background as a gaussian blured image
When both Morph opening (1.) and Minima (2.) checkboxes are unchecked im_browser estimates background as a gaussian blured image. This is the easiest method from the computational point of view but may give rise of artifacts when the strel size is too small.
How it works:
1. Generate background image by filtering of the image with a large Size parameter of the gaussian filter (6.).
Background to be subtracted, obtained with Size=120, Sigma=40.
2. Subtract the background image from the current image.
Result of background subtraction
3. Show profiles for the specified line, if 7. is checked.
Back to Index --> User Guide --> Panels