Metallurgical Image Analysis Software System
I. Technical features
* 32bit WINDOWS (Chinese version) operating platform
Software adopts Chinese menu and standard WINDOWS interface and is applicable to current leadingPC operating system such as WINDOWS95 & 98/NT (Chinese version) etc. Use practice of material workers is specially considered. Even if they do not have expertise in computer and digital image, they can use it freely after only a few days.
* Real-time acquisition, storage, processing and printing of metallurgical image.
II. Image acquisition
The system acquires images real time at the speed of 30 frames/sec. Image is shown on monitor
synchronously for easy measurement analysis, reduction of fatigue caused by eye lens observation and for even easier observation and analysis by many people at the same time. Image resolution is 768X576 and maximum gray level is 256.
1. Image activation: microscopic image is real-time shown on the computer screen. Adjust objectivetable to select optimal field of view and adjust microscope focus to make image clearer. At the moment, the image is in active condition, i.e. 25 images are shown each second.
2. Image acquisition: acquire current microscopic image into the memory and image will not bereal-time displayed on the computer screen. At this moment, image is in frozen condition and can be processed.
3. Saving image: select appropriate path in file dialog box and enter correct file name to save currentmicroscopic image to disk for file in the format of 256 gray scale BPM bitmap.
4. Opening image: select the file to be opened in the file dialog box to call out image in the disk forprocessing.
5. Printing image: turn on printer or use video copying machine to print photo-level picture instantly(optional). Picture brightness, contrast, and size are adjustable, thus relieving of darkroom trouble.
III. Image processing
* Terms
1. Microscopic image: digital image obtained through acquisition, digitalization and quantization of
the sample optical metallurgical image observed in microscope by means of camera and image processor. Actually, it is quadratic function of image related to position coordinate.
2. Gray scale: bright and dark of image point, the brightest (white) is 255 and the darkest (black) is 0.
3. Gray level: number of bright and dark scales of image point, 256 for this system.
4. Image processing: improve image quality through specific calculation of digital model.
5. Image identification: pick up some features of image and differentiate different targets based on these features.
6. Image intensification: improve visual effect of image.
7. Image filtering: remove impurities on the image and highlight some detail features.
8. Image segmentation: divide image into significant target areas.
* Image intensification
1. Adjustment of brightness and contrast: move scroll bar in the dialog box or enter a number between-50 to 50, positive number increasing contrast and negative number decreasing contrast.
2. Linear transformation of gray scale: change the shape of gray scale curve to change the range of input and output gray scale. If input gray scale range is narrow and output range is wide, brightness and contrast can be increased; if input gray scale range is wide and output range is narrow, image detail of a certain gray scale section can be highlighted.
3. Nonlinear transformation of gray scale: change the shape of gray scale curve to increase brightness and contrast and highlight the detail of gray scale image.
4. Balanced transformation of histogram: this function can increase the image contrast and highlight the detail of gray scale image.
5. Linear stretch transformation of histogram: this function can increase the image contrast and highlight the detail of gray scale image.
* Display of histogram
To display gray scale histogram of current image in dialog box and probability distribution of image elements with different gray scale values in the whole image.
* Filtering
1. Smoothed filtering: smoothed filtering can remove or weaken such noises as stain, water spot,scratch, cavity in the image and effectively improve image quality.
2. Sharpened filtering: the purpose of sharpened filtering is to intensify such details as grain boundary,phase boundary and inclusion. Different filter operators are effective for different images.
* Image segmentation
The purpose of image segmentation is to separate characteristic object from background and to transform original gray scale image into black and white image for easy processing. For example, separating graphite from matrix for ductile iron, graphite represented by black and matrix by white. Image segmentation is the key step for image identification.
* Image inverse
The purpose of image inverse is to inverse black and white of gray scale image.
* Image convolution
Select different convolution kernels and use convolution operation to conduct smoothing, sharpening and edge enhancement processing for images. For relevant theory, refer to documents concerning image processing.
* Tow-value image processing
Functional description of respective processing methods
1. Corrosion: peel off a circle around the characteristic object and use this function to remove small noise granule and pitting.
2. Expansion: expand a circle around the characteristic object and use this function to fill small cavities.
3. Open: conduct corrosion operation and then expansion operations of the same frequencies. Use this function to realize separation of contiguous granules.
4. Close: conduct expansion and then corrosion operations of the same frequencies. Use this function to realize conjunction of separate characteristic objects.
5. Pick-up of grain boundary: pick up boundary image of granule.
6. Pick-up of framework: pick up framework of granule, e.g. pick up grain boundary as single line width.
* Image edit:
1. Window setup: in default condition, the image processed and edited is the whole image. To process the image of the whole area, it is necessary to set up image window.
2. Window cancellation: to set up new window, first select “window cancellation” andthen “window setup” to re-set up the object of image processing to the whole image area.
3. Amplification: amplify the currently selected window.
4. Minification: minify the currently selected window.
5. Image superposition: for arbitrary drawing.
6. Word superposition: for arbitrary entering of words.
7. Erase: move eraser to erase image.
8. Positioning: position and gray scale value of image element point currently at cursor are displayed in the status bar.
* Granule analysis
1. Description of granule parameter:
Xs, Ys Coordinate of upper left hand corner of granule
circumscribed rectangle
Xe, Ye Coordinate of upper right hand corner of granule
circumscribed rectangle
Xc, Yc Granule centroid
Xw Granule width
Yw Granule height
Area Number of granule image elements
Perimeter Number of image elements at the periphery of granule
D0 Feret diameter in X direction of granule
D90 Feret diameter in Y direction of granule
D45 Feret diameter in 45°direction of granule
D135 Feret diameter in 135°direction of granule
Euler number Euler number of granule
Number of holes Number of holes contained in granule
Roundness One of granule shape parameters
Concavity One of granule shape parameters
Diameter of circle area Diameter of circle of the same area as granule area
2.Granule screening: depending on different screening principles, this function can cancel or retainselected granules. Granule screening is carried out on basis of granule size, i.e. image element.
3.Filling cavity: select the maximum area of the cavity to be filled. This function can fill the cavitywhose size is less than the specified value.
4.Calculation of granule parameter: systematically calculate parameter of granules and output theresult to the dialog box of report file.
5.Granule counting: systematically count parameter of granules and output the result to the dialogbox of report file.The above items can be saved and printed.
* Measurement
1.Setting up scale: before measurement, set up scale to confirm current amplification, or measuring resultmay be inaccurate.
2. Length: use the points at the upper left and corner and low right hand corner of mouse rectangle formeasurement. True length of straight line is real-time displayed in status bar.
3. Angle: use mouse to move the top and two laterals of the corner and angle is real-time displayed instatus bar.
4. Rectangular area: use the points at the upper left and corner and low right hand corner of mouserectangle to minify rectangle and rectangle area is real-time displayed in status bar.
5. Circular area: at the upper left hand corner (low right hand corner) of circumscribed rectangle, usemouse to move ellipse and ellipse area is real-time displayed in status bar.
* Special metallurgical software
1. Grading of average grain size
2. Grading of carbon steel graphitization
3. Measurement of phase content
4. Grading of pearlite graphitization
5. Grading of ductile iron nodularity
6. Measurement of cast iron pearlite
7. Grading of nonmetal inclusion contained in steel
8. Depth measurement of steel carburized layer
9. Grading of lumpy carbide of high speed steel
10. Length measurement and grading of gray cast iron graphite
11. Grading of ferrite grain size mixed with pearlite
12. Grading of graphite carbon in steel
13. Measurement and grading of gray cast iron phosphide eutectic
14. Inspection and grading of carbon steel graphitization (standard of Ministry of Electricity)
15. Needle length measurement of medium-carbon steel martensite
16. Thickness measurement of decarburized layer of steel
17. Measurement of carbon fiber porosity
18. Spacing measurement of gray cast iron pearlite strip
19. Measurement of aluminum coating
20. Size grading of ductile iron graphite
21. Quantity measurement of gray cast iron carbide
22. Vermicular rate measurement of vermicular graphite iron
23. Quantity grading of vermicular graphite iron phosphide eutectic
24. Quantity grading of vermicular graphite iron carbide
25. Measurement of ductile iron carbide
26. Grading of cast iron high manganese steel inclusion
27. Metallurgical examination of cuprous oxide of precious metal materials containing copper
28. Grain size of cast aluminum copper
29. Grade measurement of automobile carburized gear martensite
30. Fn test of zirconium alloy tubular goods
For detailed specific operating method and precautions, refer to “User’s Manual of Wantai Metallurgical Image Analysis System” (ten sets of software are selected for each analysis system depending on the need of different users).
* Programming guideline
VMS supplied subprograms are in library file VMS.LIB. They are compiled in C language for easy calling and second development by users.
For specific operating methods, refer to ““User’s Manual of Wantai Metallurgical Image Analysis System”