University:
Boston University
Course:
CAS CS 501 | Computer Science Practicum
Academic year:

2020
Views:

158

Pages:

18
Author:

Beuningen1c1q

https://en.wikipedia.org/wiki/Bilinear_interpolation Upscaling & interpolation  None: Copies pixels into neighboring spots.  Will look “blocky” with large amounts of upscaling.  Bilinear: Fits straight lines through 4 neighboring pixels and interpolates  Bicubic: Fits a 3rd order polynomial to 16 neighboring pixels and interpolates Adds new points: upsampling!  Open 5x5.tif and let’s try this out. 5x5 8-bit 25x25 8-bit No interpolation 25x25 8-bit bilinear 25x25 8-bit bicubic      Zoom into 5x5.tif with Ctrl-+ With the box tool, select the middle row of pixels. Choose AnalyzePlot Profile or Ctrl-k to see a profile plot of the middle row. Click in the image to undo the selection (or Ctrl-Shift-A) Open the resize tool with Ctrl-E. Increase to 25x25 pixels, pick an interpolation method, add a box, and plot its profile. Profiles for each resized image  The profile data was extracted for each resized image using the List button in the lower left of the Plot window.  It is important to understand what image algorithms are doing mathematically to your image before applying them! 8-bit images  Open another image file, Citgo_Sign.tif  Hover your mouse over part of the image. The x-y location and pixel values for the RGB channels will be displayed in the lower let of the main window.  This is a 8-bit image so the value range is 0-255  When displayed, every grayscale image has a colormap as you’re using an RGB display!  Display the colormap: ImageColor  Display LUT Direct 8-bit cool LUT Sepia LUT  Green Blue Fire LUT Choose some other LUTs: Remember this just effects the display, not the image on disk which remains an 8-bit grayscale image. Color Images  ImageJ supports two types of color images, RGB and composite.  RGB images are standard color images.  E.g. pictures from your smartphone or a three-color CCD or CMOS detector  Composite images are sets of grayscale images that are assigned colormaps and then combined to form an RGB image. RGB image Stacks and Hyperstacks  ImageJ handles 4 dimensional images using a stack.  The 4th dimension can be time or wavelength channel.  Example 1: A set of grayscale microscope images taken over several minutes (grayscale video)  Example 2: A set of grayscale images acquired with different wavelength filters and a filter wheel.  Example 3: An RGB image  A hyperstack adds a 5th dimension  Example 1: A set of grayscale images acquired with different wavelength filters with multiple time steps per wavelength (multispectral video).  Example 2: A set of RGB images acquired over time (color video) RGB images  Open the image tetraspeck.tif. These are ThermoFisher fluorescent microspheres.  Convert this from an RGB image to a 3-channel stack:  ImageColorsSplit Channels  ImageStacksImages to Stack  The scrollbar switches between slices in the stack.  Press Ctrl-Shift-C to open the B&C control. Each slice gets the same B&C setting.  Hold the Shift key when moving between slices – this will do the Auto B&C command for each slice. Back to RGB  Convert the stack back to a set of images:  ImageStacksStack to Images  Click on each image and choose the matching LUT (Red for the red image, etc.)  Re-create the RGB image:  ImagesColorMerge Channels  Check Keep source images  The RGB image will open alongside the other 3. Composite Image  Open Rat_Hippocampal_Neuron.tif  This is a Fiji sample image  This is a 4-channel fluorescent image with a white light image.  The per-channel LUT information is stored by ImageJ in the TIFF file header.  Choose ImageColorChannels Tool  Choose Composite and try the options.  Click More – try creating an RGB from your composite.  Now choose Color, then click More.  Here you can change the LUT per channel Ctrl-Shift-Z Open channels tool So Far…  Image Format Types  Vector Images  Raster Images  Raster Image Quality  Color  Digital vs. Print  RGB and CMYK  Getting Started with ImageJ  Opening and saving images  Selections and cropping  Image Adjustments  Measurements      Area and dimensions Histograms ROI and masking Thresholding Quantification of intensity  Macros  Recording and running  Writing your own Macros  ImageJ macros can be saved, edited, assigned to keys, and shared amongst friends.  The macro language is straightforward and supports all the built-in ImageJ commands  This gives users the ability to create re-usable image processing programs without having to learn a language like R, Python, or Java.  Choose PluginsMacroRecord  The window that opens will now record your commands.  Leave it open! Measuring Images  Open banana.png  Click the straight line tool, then draw a line between some of the centimeter marks on the ruler.  Choose AnalyzeSet Scale.  Set the Known distance to the cm’s you choose.  Set the Unit of Length to cm  Press Ctrl-Shift-A to remove the line. Measurements cont’d  This is an RGB image, choose ImageType8-bit to convert it.  The background is uneven. ProcessSubtract Background and enter 250 to do a rolling ball subtraction.  Ctrl-Shift-C to adjust B&C if you like.  Draw a rectangle around the banana and choose ImageCrop Measurements cont’d  Thresholding: Pick a pixel value V.   Pixels >= V  assign max value Pixels < V  assign zero value.  Using a statistical threshold is much more reproducible than selecting thresholds by hand.  ImageAdjustThreshold Click Dark background. Select a few, see what happens.  Choose Otsu and click Apply. Close the Threshold window. Red indicates value below the threshold. Measurements cont’d  Fill in the holes in the banana with ProcessMorphologyGray Morphology  Apply a close with a radius of 5.5 pixels.  The close applies a dilation followed by an erosion.  Now use AnalyzeAnalyze Particles  Under Size enter 1  Check Display Results and Summarize  Choose Outlines under Show.  More measurements can be set: AnalyzeSet Measurements Macro window  It should look something like this:  Close all of your windows except the macro recorder.  Click the Create button. The full macro editor will open.  Close the macro recorder.  Click the Run button. What happens?  The macro editor can save your macro, and you can run it any time.  This allows the exact same process to be applied to multiple images.  ProcessBatchMacro will apply a macro to a directory of images.

Introduction to ImageJ Part 3

Related Documents

Recommended Documents

Description

Related Documents

Free up your schedule!

Take 5 seconds to unlock