EN_OPUS6.0 Object.book
Contents
1. OPUS Spectroscopy Software User Manual OBJECT gt lt BRUKER LL O 2006 BRUKER OPTIK GmbH Rudolf Plank Str 27 D 76275 Ettlingen www brukeroptics com All rights reserved No part of this publication may be reproduced or transmitted in any form or by any means including printing photocopying microfilm electronic systems etc without our prior written permission Brand names registered trade marks etc used in this manual even if not explicitly marked as such are not to be considered unprotected by trademarks law They are the property of their respec tive owner The following publication has been worked out with utmost care However Bruker Optik GmbH does not accept any liability for the correctness of the information Bruker Optik GmbH reserves the right to make changes to the products described in this manual without notice This manual is the original documentation for the OPUS spectroscopic software Table of Contents 1 UO a LAE 1 2 Technical Requirements css esses esa ea sex aa aaa 3 2 1 Soltwate EA UNA AAA 3 2 2 Hardware Requirements zta sr a0 Meron ar 3 3 Image Processing Functions cece cece cece cece esco 5 3 1 Compute Grayscale Mi ds bu bage Gi ei Mt Gags 3 a2 Media a hse Mune ood eee Rees aaa ae antes aaa 6 33 EEE e E ie a reas SEA Sonat CANO CA keene NE 6 3 4 Stretch Histogram e Ed A O E ees 7 3 3 Bi nariz tion A Glade a VANE CREE ORE RANE eked A 7 3 6 BLOG sa Bae E Ee ega2. 3 Negative This function inverts the intensities of an image and thus produces the negative of the original grayscale image Use this function if the objects of interest are darker than the background because the algorithm for detecting objects assumes that the objects are lighter than the background Figure 3 Negative Image of the Sample Bruker Optik GmbH OPUS OBJECT 6 Stretch Histogram 3 4 Stretch Histogram If the pixel intensities in an image do not cover the full dynamic range from 0 to 255 e g from 80 to 255 the negative image will be too dark to see details So if the total contrast of the image is too low you can use the Stretch Histogram function to change the intensity distribution in such a way that the whole dynamic range is covered This operation is only necessary if you want to see details in the image by visual inspection If you work with automatic object detection this step can be omitted 34600 34800 35000 35200 35400 35600 35800 i N KO N EE amp Q o N o B N 8 Ea E o N N o E N a GO N Q B be N 8 o N Figure 4 Result after using Stretch Histogram 3 5 Binarization This function assigns to all image pixels of which the intensity is above a cer tain threshold the maximum intensity 255 white while those pixels with an intensity below a certain threshold are assigned to the minimum intensity 0 black The threshold is set int
3. 0 23800 24000 24200 24400 24600 24800 25000 25200 Figure 6 Image after Binarization 3 6 Erode This function can only be applied to a binarized image It eliminates small objects by moving a 5x5 rectangle over the image and deleting the pixel at the central point of the rectangle if there is at least one black pixel inside the 5x5 rectangle All regions consisting of less than 25 pixels are deleted in this way Thin connections between larger objects are also erased and the size of all objects is reduced by at least 2 pixels on each boundary 34600 34800 35000 35200 35400 35600 35800 23400 23600 23800 24000 24200 24400 24600 24800 25000 25200 Figure 7 Image after Erosion Bruker Optik GmbH OPUS OBJECT 9 Image Processing Functions 3 7 Dilate This function is the counterpart to the Erode function It can also be applied to binarized images only The function produces just the opposite effect by mov ing a 5x5 rectangle over the image Wherever the central pixel of the rectangle is placed over a white object pixel all 5x5 surrounding pixels are converted into white pixels By doing so small holes are filled and the white regions are re enlarged to their original size 34600 34800 35000 35200 35400 35600 35800 23400 23600 23800 24000 24200 24400 24600 24800 25000 25200 Figure 8 Image after Dilation 3 8 Opening Normally Erode is followed by Dilate in order to retain the size of the remain ing o
4. 400 24600 24800 25000 25200 Figure 12 After the Measurement Bruker Optik GmbH OPUS OBJECT 13 Image Processing Functions Bruker Optik GmbH OPUS OBJECT 14 Index Numerics 3D file 12 B Background measurement 12 Binarization 7 8 Binarized image 3 8 10 11 BMP format 4 D Dilate 3 8 10 E Erode 3 8 9 10 Erosion 9 10 F False color 11 Find regions 3 11 Frame 3 Frame grabber board 3 G Grayscale image 5 6 8 H Histogram 7 8 I Intensity 6 7 8 Intensity distribution 7 8 J JPEG format 3 4 M Measure marked positions 12 Median 6 N Negative 6 Negative image 7 O Opening 10 Overview image 1 3 Overview window 5 R RAM 3 RGB image 3 5 RGB overview image 3 5 S Sample measurement 12
5. age capacity of 46 Megabyte These image sizes can still be han dled if RAM has the recommended disk space of 256 Megabyte To avoid swapping do not use a RAM with a lower storage capacity The software enables you to store the images in JPEG format which com presses the data significantly and thus reduces the necessary disk space How ever this compression is not loss free i e it slightly modifies the data and thus cannot be used as input images for operations like Erode Dilate or Find Regions which require an exact binarized image Therefore even if you select Bruker Optik GmbH OPUS OBJECT 3 Technical Requirements JPEG format the software automatically switches to BMP format if required Note that the compression does not reduce the disk space needed in RAM because JPEG images are decompressed to BMP format and require the same disk space in RAM as BMP images As the images require a high storage capacity the available free disk space should be as large as possible In addition the possibility of saving data on a server or a CD writer should be taken into consideration right from the begin ning Finally storing reading and computing such large amounts of data requires a fast hard disk and powerful CPU Bruker Optik GmbH OPUS OBJECT 4 Compute Grayscale Image Image Processing Functions The Image processing functions can be accessed via the pop up menu of the overview window provided you have registere
6. apacity requirements If you scan a sample area of lcm using the IRScopell or Hyperion microscope with 4x VIS objective but without an optical adapter one image frame covers an area of 0 848mm in x and 0 636mm in y The generated overview image is composed of 192 individual image frames 12 frames in x direction and 16 frames in y direction with each frame having a size of 1 3 Megabyte That means that the resulting RGB overview image requires a storage capacity of 250 Megabyte If you use the 15x IR objective instead of the 4x objective one image frame covers only an area of 0 29mm in x direction and 0 217mm in y direction The resulting overview image is then composed of 1 610 image frames 35 frames in x direction 46 frames in y direction requiring a storage capacity of 2 Gigabyte During the image processing two grayscale images are computed and stored in addition to the original RGB image They both increase the amount of image information by about 70 of the original RGB image As images of these sizes cannot be stored in RAM the software works disk interactively it is assumed that an image having the same size in x direction as the overview image and consisting of one frame height in y direction can be stored in RAM In case of an image generated by the 4x VIS objective this means that 12 image frames 12 x 1 3 15 6 Megabyte must be stored in RAM while in case of the 15x IR objective this would be 35 image frames requiring a stor
7. arked by circles In addi tion the regions representing the objects formerly white regions are combined and represented in false color with the underlying image Bruker Optik GmbH OPUS OBJECT 11 Image Processing Functions 34600 34800 35000 35200 35400 35600 35800 23400 23600 23800 24000 24200 24400 24600 24800 25000 25200 Figure 11 Found Objects for Measurement 3 10 Measuring Spectra of the Found Regions As long as the video image window is open all found regions can be measured automatically using the Measure Marked Positions function of the Video menu This measurement function works in the same way as if the regions were defined interactively Before starting the sample measurement a background spectrum must be acquired Especially if a larger number of spots has to be measured it is recommended to repeat the background measurement after every n sample measurement For this purpose you have to define a x y stage posi tion for the background spectrum The measurement can be stopped by clicking on the Stop button The circles around those spots which have been measured adopt the color of the corresponding spectrum and the letter M changes to the running number of the spectrum in the 3D file A tick mark indicates that the spot has been measured Bruker Optik GmbH OPUS OBJECT 12 Measuring Spectra of the Found Regions 34600 34800 35000 35200 35400 35600 35800 23400 23600 23800 24000 24200 24
8. bjects The combination of these two operations is called Opening first it completely eliminates all small objects by the erosion and then it fills small gaps and re enlarges the remaining objects to their original size The effects of an Erode and Dilate operation become clearer if the image is zoomed until indi vidual pixels become visible The following figure illustrates the effects of an Opening operation Bruker Optik GmbH OPUS OBJECT 10 Find Regions 34660 34680 34700 34720 34740 34760 34780 34800 34860 34880 34700 34720 34740 34760 34780 34800 34060 34680 34700 34720 34740 34760 34780 34800 Figure 9 Effect of an Opening operation 3 9 Find Regions This function searches the binarized image for white areas and determines their position center of gravity and size Typically most samples contain a large number of tiny objects with the size of a few pixels In most cases these tiny objects are not of interest as they may represent undesired information like noise or dust particles After having found all regions a dialog box pops up which allows the user to discard all objects smaller than a user defined mini mum size Total no of objects found 202 Max object size micron 1674 30 Min object size micron 5 29 Discard objects smaller than micron m Number of objects above threshold 131 Abbrechen Figure 10 Image Statistics and Defining a Threshold If you click on the OK button the found regions are m
9. d your OPUS OBJECT package See OPUS Reference Manual for details Mouse Mode gt Video Image gt Measurement Spots Grid gt Starting Measurement b Export A Import gt Delete A Compute Grayscale Image Median Negative Stretch Histogram Binarization Erode Dilate Find Regions View RGB Overview View Current Result Restore Original Grayscale Figure 1 Image Processing Pop up Menu 3 1 Compute Grayscale Image First a grayscale image is created using the Compute Grayscale Image func tion After the computation is completed the grayscale image instead of the original RGB image is displayed You can switch between the RGB Overview image and the result of the last image processing operation using the View RGB Overview and View Current Result functions Bruker Optik GmbH OPUS OBJECT 5 Image Processing Functions 34600 34800 35000 35200 35400 35600 35800 23400 23600 23800 24000 24200 24400 24600 24800 25000 25200 Figure 2 Grayscale Image of the Sample 3 2 Median This function eliminates extreme intensity values at certain pixels by moving a 5x5 mask over the image sorting the corresponding 25 intensity values in a lin ear array selecting the EA intensity value and substituting the original pixel at the center of the 5x5 mask by this value Use this function if a few pixels being surrounded by normal pixels are completely wrong and have to be eliminated 3
10. eractively in the following window which opens after Binarization is started Bruker Optik GmbH OPUS OBJECT 7 Image Processing Functions Pick Threshold Value 34600 34800 35000 35200 35400 3 34600 34800 35000 35200 35400 3 200 24400 24600 24800 25000 25200 200 24400 24600 24800 25000 25200 Figure 5 Binarization Window Interactive Threshold Setting At the bottom of the window the image histogram i e the intensity distribution of the image is shown The x axis lists all possible intensity values of the gray scale image from 0 to 255 the y axis displays the number of pixels having a certain intensity value x The aim of the binarization is to show all interesting objects in white on a black background This is only possible if the objects are lighter than all other areas of the image Ideally the histogram has two peaks the peak at lower intensities represents the dark background while the peak at higher intensities arises from the lighter objects To separate the objects from the background you have to define an intensity value between both peaks This can be done by moving the vertical line with the mouse to the right or to the left and watching the effect on the right hand image The binarized image can be further processed using the Erode and Dilate functions to remove areas that are too small Bruker Optik GmbH OPUS OBJECT 8 Erode 34600 34800 35000 35200 35400 35600 35800 a f NM KO 23400 2360
11. n Ede 9 3 7 LAI REA 10 3 8 OPE a gata ag atata ip iy did 10 3 9 FIORE ad ESA A ERE 11 3 10 Measuring Spectra of the Found Regions u z 12 1 Introduction The OPUS OBJECT package contains a set of image processing functions for detecting objects within digital images This software can be best used in con nection with the OPUS VIDEO software and an x y stage controlled by the OPUS MAP software Using the OPUS MAP software you can interactively define an interesting sample area Then the OPUS VIDEO software automati cally generates an overview image of that sample area which is assembled from various single image frames at different stage positions The OPUS OBJECT package can then be used to find objects of interest in the overview image to extract their size and position and to measure IR spectra of the found objects by moving the x y stage to corresponding position Bruker Optik GmbH OPUS OBJECT 1 Introduction Bruker Optik GmbH OPUS OBJECT Software Requirements 2 Technical Requirements 2 1 Software Requirements The OPUS OBJECT software requires the OPUS VIDEO software to control the frame grabber board and to produce the VIS images and the OPUS MAP software to control the computer controlled x y stage 2 2 Hardware Requirements The analysis of large images requires a high RAM and hard disk storage capac ity depending on the size of the overview image area The following example illustrates the storage c
Download Pdf Manuals
Related Search