tnimage Help File ------------------------------------------------------------------------ This help file is is ordinary ASCII format. New comments or corrections can be added as needed, provided that the following format is adhered to: (1) Line length can not exceed 80 characters. (2) The first character of each line is either a digit (indicating the help topic) or a space. (3) A maximum of 40 lines per help topic. No copyright restrictions apply to this file. If you pass this file to another person, please indicate in the file that you have modified it, to avoid confusion. ------------------------------------------------------------------------ 1 General Help - (F1) Unless you select a region, most operations will be applied to the entire image. To select a region, click on one corner and drag to the other corner with the mouse. To add text - move to the desired location and begin typing. To move the image around - Click on one of the 4 arrows in the menu bar. The amount of movement can be changed by pressing the gray (+) or (-) keys. These keys also change the distance the cursor will travel when it is moved with the arrow keys. There are also shortcut methods for some operations: Selecting foreground color - click on desired color in the displayed palette with the left mouse button. Selecting background color - click on desired color in the displayed palette with the right mouse button. Drawing lines - Press F2 to toggle line drawing mode on/off. Selecting areas - Click & drag or press F3 for non-rectangular region. Clicking on the menu bar also turns off line drawing, box drawing, densitometry, region scanning, and most other operations. If image colors are wrong after some operation, try pressing Alt-R to rebuild the screen display. If program appears stuck, click Cancel. The Main Cancel Button at left always cancels the current operation. Some windows cannot be dismissed until the operation has been cancelled. If Dismiss button seems not to work, try clicking Main Cancel Button. See the manual tnimage.ps for more detailed information. 2 Help - Pixel interaction mode Determines how 2 pixels will interact when an image is copied over another or loaded from disk. You can subtract 2 images, or offset the same image by a small amount and then subtract it (to create a 3D effect). Overwrite ( default ) - This is the normal, fast mode. Maximum - Pixel is set to the larger of the 2 values. Minimum - Pixel is set to the smaller of the 2 values. Add - The 2 values are added. Subtract - The 2 values are subtracted. XOR - The 2 values are XOR'd. Average - The 2 values are averaged. For the pixel values to become permanent, you must create a new image by selecting "Create image" from the 'Files' menu. Otherwise, they will revert back to their original values when the image is redrawn. 3 Help - Save image Transfers an image or selected region to disk. Save entire image - This option will cause the currently-selected image to be saved in its entirety. Save selected region - Causes the selected screen region (which may contain parts of 1 or more images) to be saved. If multiple images overlap, the pixels from the foreground image are saved. TIFF/PCX/GIF/IMG/Custom - Selects the image file format. Color/gray scale vs. monochrome - Select monochrome for black-and-white images. Monochrome images will be much smaller than color/gray scale images but contain no color or gray. Color and gray scale images are treated identically. Filename - Click on this box, and type in the desired filename, then press or click the mouse when finished. Image number - Not normally changed. This indicates which image is currently selected in case you want to save the entire image. Extra TIF param.(for `Other') - Selects non-standard color depths. Only used if file format is TIF or Custom and Image Type is `Other'. 4 Help - Print Prints an image or selected region on a PostScript or PCL printer. These settings apply to all printers: Positive/negative - Positive prints the image exactly as it appears on your screen, negative prints the inverse. Grayscale, RGB, CMY, CMYK - Print mode. Grayscale - Ignores colormap, sends image as grayscale. This is the recommended setting for grayscale images unless the grayscale mapping has been changed. RGB/Indexed color - Uses the image's colormap if present. This is the recommended setting for sending 8-bit color images such as GIFs to PostScript printers. Not recommended for PCL printers. CMY - Uses the image's colormap if present. This is the recommended setting for sending all color images to any printer. CMYK - Recommended for high-end printers. Gives a wider range of colors. Print entire image/ print selected region - Select whether you wish to print the entire image, or only the screen region which you selected most recently (if no region was selected, it will print the entire image). Printer command - enter a command to be used for network printers. Print device/file name - the printer name, e.g. /dev/lp0 or a filename if you are printing to a file. These settings apply only to PCL printers: Resolution - Enter a value such as 150, 300, 600, or 1200 dots/inch, depending on the resolution of your Laserjet-compatible printer. Dither size - This determines the number of gray or color levels which appear in the print-out. For example, dither size of 8x8 corresponds to 64 different possible levels. A larger dither size also means the print-out becomes larger. Print quality, Media type, Depletion - Give fine control over the printout. These settings apply only to PostScript printers: Horiz. image size - the width of the printout in inches. Image aspect - Greater than 1 elongates the printout vertically, less than 1 stretches it horizontally. Rotation - degrees counterclockwise. Because the center of rotation is the lower left corner of the image, it may be necessary to increase the horizontal offset to keep the image on the page. Paper size - This should be set to match the paper tray on your printer. 5 Help - Change palette Changes the color values for 8-bit images. Click on the desired palette. Attaching the palette to the image causes the change to be permanent. These pre-defined palettes are available: Gray scale Spectrum Multi-color 1,2, or 3 RGBI Black to green Zebra Other (selects from 10,000 other pre-defined palettes). Brightness - changes the color saturation. Rotate palette - causes the color map to be shifted to lower or higher pixel values (depending on where you click). Create - Allows you to graphically create your own palette. The palette can also be saved on disk. 6 Help - Densitometry Measures the density and area of an object on the screen. Auto background - uses fuzzy k-means to automatically determine foreground and background pixel densities. Pixel compensation - If set to "on", the pixels are converted into values proportional to the optical density originally calculated by the scanner. If set to "off", the results are raw intensity values. This only has an effect if the image contains an optical density table. Maximum signal - Selects whether the image is a white-on-black or black-on-white image. If the maximum signal is 'black', then a completely black region will have the strongest signal. Area selection - Automatic mode: Click near the center of the region to be analyzed. The boundaries of the region will be calculated automatically. The area actually analyzed temporarily changes color. You may need to adjust the background level so the correct area is analyzed. For extremely noisy images, it may be helpful to delineate the outer boundary of the region to be analyzed by first drawing a black line around it. Manual mode: You must select a rectangular region each time. Fixed size mode: You must select a rectangular region for the first measurement. This is used as a template so that all subsequent measurements use the same number of pixels. See also `Scan region...' for an alternative method of densitometry. 7 Help - Filter - Filters an image or selected region. Low pass - Blurring filter High pass - Sharpening filter Laplace - Edge detection Sobel - Sobel edge enhancement Background subtract - Removes only low spatial frequencies. Eliminates slow gradients and broad smudges from image. The entire image will probably get lighter or darker, depending on whether the "maximum signal" is set to "white" or "black". Background flatten (de-trending) - Removes color gradients. Noise filter - Eliminates extraneous pixels. Sharpen / - Adds shadow at 45ø angle (freeze-fracture effect). Sharpen | - Adds shadow at 90ø angle (freeze-fracture effect). Sharpen / - Adds shadow at 0ø angle (freeze-fracture effect). Kernel - Determines the number of surrounding pixels to be convoluted with each pixel. In conjunction with kernel multiplier, this determines the size range that will be emphasized or smoothed. Kernel multiplier - Determines the area that is covered by the selected kernel. A higher value results in filtering lower spatial frequencies. Range (median) - Determines the maximum deviation from the median value that is permissible during noise filtering. If a pixel deviates from the median by more than this amount, the median is substitued. Amount of filtering - How much filtering is desired (10=maximum). Pseudo color images (such as GIFs) should be converted to 24 bits/pixel before filtering. 8 Help - Scan region Performs densitometric analysis on a screen region by scanning it from one side to the other. Select coordinates mode - In this mode, you must reselect the region to be scanned each time. This is done by: Click on one starting corner with left mouse button. Drag to the other starting corner and release mouse button. Click on one ending corner (This can be on the left, right, above, or below the starting corners). Drag to the other ending corner and release. Scanning will then start in the direction of 'start' to 'end'. Maximum (black or white) - Determines whether black or white is to be regarded as the largest signal. Rhomboid, 90°(4 pts.): the starting and ending edges will automatically 'snap' to the closest value (0 or 90 degrees), causing diagonal areas to be scanned as a series of vertical or horizontal lines. Select 4 points and press any key to start. To select next area, select 4 more points. Click 'cancel' when finished with densitometry. Rhomboid (4 pts.): starting edge can be at any angle. Fixed width: Scans a rectangle of constant width. Select 2 points, one at each end of region to scan. Press any key to start measurement. To select next area, select 4 more points. Click 'cancel' when finished with densitometry. Automatically save scan - If checked, the data will be saved in an ASCII file after each scan. Filename - Filename under which the scan will be saved. If the 1st 8 characters are all digits, the filename will be incremented as a number each time, e.g., '1000.dat' will become '1001.dat'. Peak area measurement - Use mouse to click and drag desired region on the density profile plot. The area is shown in the lower right. Or, click on "Peak areas" to get a table of all peaks. You can save this table by clicking on `save to disk'. Manual Baseline - click on the graph to create control points for a Bezier curve for the baseline, then click on "Finished" or press to subtract the baseline from the graph. 9 Help - Fill an area Flood-fills a screen area with color. Flood fill will not go outside the selected area. Fill type - Select between a solid color, or a gradient which changes in the vertical or horizontal direction. Fill color - The color to be used when filling with a solid color. Maximum gradient color - The color value to be used at the top (for vertical fills) or far left (for horizontal fills). Image should be at least 24 bits/pixel for good results. The 'maximum' color can be higher or lower than the 'minimum' color. Minimum gradient color - The pixel value to be used at the bottom (for vertical fills) or far right (for horizontal fills). Interactive filling mode - click on a point inside the region to fill. Repeat for each area to fill. At any time, you can stop the fill process by pressing or clicking on Cancel button. If no filling occurs, it may be necessary to change the border colors. Max border color _ The region that includes all contiguous pixels between Min border color these two values will be filled. Preselect filling mode - Select desired area to fill in advance by dragging a rectangle or double-clicking on an object. 10 Help - Click box The minimum and maximum acceptable values are shown. Click on the arrows on each side of the indicator bar to increase or decrease the value, or click and drag the vertical position indicator until the desired value is obtained. You can also type the number directly, by clicking on the number field. 11 Help - Read image Transfers an image from disk to the screen. Filename selection: Click on the "Filename" box, then type the filename, or click again to select a file from the list. Click on desired file with the mouse. The TAB key toggles between the filename list, directory list, and drive list. Clicking on the desired list has the same effect. Click on the small triangles in the upper and lower right of the list to scroll through the list. Or, click on and drag the indicator to move rapidly through the list. The arrow keys can also be used to scroll up and down in the currently-active list. To change drives, click on the desired drive name. Positive/negative: Determines whether the image should be color-inverted. Vertical/horizontal: Determines the screen orientation for the image. X-position: Left x-coordinate of the image. Y-position: Upper y-coordinate of the image. X size: Final size of the image in the x dimension, as a fraction of the original size (0-1). (The x or y dimension is relative to the orientation of the image, not the screen). Y size: Final size of the image in the y dimension, as a fraction of the original size (0-1). Note: Targa format images must have the extension .TGA to be readable. 12 Help - Create Palette Creates a new, custom palette for 8-bit images. To change Red, move the mouse cursor to the top box. To change Green, move the mouse cursor to the middle box. To change Blue, move the mouse cursor to the lower box. Then click and drag to create a graph of the intensity of that color for each color value between 0 and 255. Click on 'OK' when finished. Click on 'smooth' to make the new palette smoother. 13 Help - Plot Click and drag on the graph to select a range of X-values to be integrated. Click OK when finished. BL Smoothness: Adjusts the frequency cutoff of the automatic baseline calculation. A higher value increases the smoothness of the calculated baseline. 14 Help - Color Intensity If the region you selected contains a monochrome region, adjust the desired change in brightness from -255 to 255 (0=no change). The pixel values in 8-bit regions will be incremented or decremented by this amount. Usually this means the image gets lighter or darker; but the actual effect on the appearance of the image is determined by the current palette. If the region you selected contains a color (16,24, or 32 bit/pixel) region, adjust the change in brightness of the red, green, and blue components as desired. This will adjust the colors in regions containing color images. To individually adjust the RGB components of an 8 bit/pixel image, you must convert the image to color first (select `Image'...`Convert color depth'). Intensity and contrast can also be adjusted interactively by dragging on the palette. "Color/brightness" tries to use the color palette if possible whereas "Change pixel value" always changes the actual image. 15 Help - Significant digits This selects the number of decimal places that will be displayed when floating-point numbers are used, such as during densitometry. 16 Help - Configure Color reduction method determines the algorithm used when viewing a hi-color image (15 or 16-bits/pixel) or true color image (24 or 32 bits/pixel) when TN-Image is in an 8 bit/pixel screen mode. This requires mapping the real colors into a palette. TN-Image can use quantization to find an optimal palette, or try to fit the image into the currently-selected palette. Luminosity factors are used when converting from color to gray scale. They should be between 0 and 1 and total to 1.0. Active Color Planes - Selects which colors (r,g, or b) to which all image manipulations are applied. Raise image on focus - If checked, clicking on an image or otherwise selecting it will bring it to the foreground. Multiple windows - If checked, each image will be created in a separate window. 17 Help - O.D. Table The optical density table is used to convert pixel values to optical density. This is also known as "gamma correction" and is only used for 8 bit/pixel images. Some image files contain their own O.D. table. TN-Image uses this information when "pixel compensation" is selected in densitometry and region-scanning modes. 18 Help - Histogram (indexed-color screen modes) Displays graphs showing the distribution of intensities and color values in the image. The top box displays the number of pixels in the selected region for each of the 256 possible pixel values (intensities). The other 3 boxes display the palette color attributes of the same pixels, for each of the 64 possible red, green, and blue palette values. Clicking on one of the the 4 plot areas selects which plot has the current focus. Displayed parameters and any area measurements, smoothing, etc., apply only to the plot that has the current focus. All 4 boxes are re-scaled whenever a change is made that affects the maximum or minimum values. 19 Help - Histogram (color screen modes) Displays graphs showing the distribution of color values in the image. The 3 boxes display the number of pixels in the selected region for each of the possible red, green, and blue color values. Clicking on one of the the 3 plot areas selects which plot has the current focus. Displayed parameters and any area measurements, smoothing, etc., apply only to the plot that has the current focus. All 3 boxes are re-scaled whenever a change is made that affects the maximum or minimum values. 20 Help - Math Performs a point-by-point mathematical transformation of the intensities or red, green, and blue values in the selected image or region. The decision of which formulas to apply depends on the color depth of the image at each point - if the point is an 8 bit pixel, the intensity is changed; otherwise, the r,g, and b values are changed. You can enter any mathematical equation using the following operators and functions: () + - * (multiplication) / (division) += -= *= /= ++ -- || && >= <= == if else | bitwise OR & bitwise AND ^ bitwise XOR ~ bitwise NOT sin cos tan sinh cosh tanh asin acos atan sqrt asinh acosh atanh abs log (natural logarithm) ln (natural logarithm) erf log10 (base 10 logarithm) exp pi cbrt (cube root) rand (random number based on specified seed) The following two-parameter functions are supported: max min pow (exponentiation) You can also use these predefined variables for intensity, red, green, blue, real, imaginary, and wavelet coefficients, respectively: i r g b re im w. Example: i=i/2 + i^2; r=g-5; g=(b+r)*g-17; b=b^1.23; Note that the r,g,and b values are bounded by 0 and 31, 63, or 255 for 15-, 16-, and 24-bit/pixel screen modes, respectively. `i' is also limited by 0 and 255. To use to data in other images, use the variables: image[ino][frame][x][y] red[ino][frame][x][y] green[ino][frame][x][y] blue[ino][frame][x][y] real[ino][x][y] imag[ino][x][y] wave[ino][x][y] where "ino", "frame", "x", and "y" are expressions for the image number, frame (if image is 3D), and x and y coordinates. Example: if(image[1][0][x][y]<121)i=244; Complex math on Fourier-transformed data is not currently implemented. 21 Help - Fourier Transform Select `forward' or `reverse' to perform Fourier transform, or select `change display only' if you only want to change which component is displayed. The display can be switched between the real and imaginary components of the Fourier transform. To perform Fourier convolution or deconvolution, you must specify two images. Both images will be FFT'd and the first image will be multiplied or divided in the frequency domain by the second image, and then reverse-transformed to give the deconvoluted or convoluted result. Don't forget to set the display to `real' before performing a reverse FFT, since the original image does not have an imaginary component and would otherwise appear black. FFT's are sensitive to the background and to the number of separate colors in the image. Improvement can sometimes be obtained by setting the screen background to black. See the manual for other tips. 22 Help - Calibration Calibrates the x and y coordinates in the image to correspond to some other parameter such as molecular weight. The calibrated values are displayed in the information window on the left. Click with the mouse at the points to use for calibration, and for each point, enter the corresponding calibration value. When finished, press any key. The calibrated value will be continuously displayed in a window. To turn off the calibration window, select "Calibration" again, and click on "Cancel". 'x units' and 'y units' both calibrate x and y coordinates to user- specified values. They are labeled 'x' and 'y' for convenience, but in fact they are independent of each other and can be in any direction. 'z units' calibrates pixel values only, and is useful when pixel value represents some other parameter such as altitude. 1-D (non-directional) - Defines a metric on the image (Recommended for morphometry). 1-D (directional) - Maps the coordinates of one dimension to a linear or logarithmic function (Recommended for gel analysis). 2-D - Maps both x and y to separate linear or logarithmic functions. 23 Help - Draw curve Linear - Fits a straight line through the control points. Bezier - Draws a Bezier curve. B-Spline - Draws a B spline curve. To draw a curve, click with the mouse at desired positions of control points. The points can be moved by dragging them to a different location. When you have enough control points, press a key to make the curve permanent. 24 Help - Creating custom format images Sets parameters for a new image format. `Target platform' is the type of computer that created the image. For most UNIX workstations, Mac's, and Cray and similar mainframes, select `Mac'. For IBM and similar mainframes, select `MVS'. `Bit packing' selects whether bits are to be packed left to right (TIF-like) or right to left (GIF-like). This is especially important if the total bits/pixel is not a multiple of 8. If `Use header' is checked, specify the file containing the header to copy, the number of bytes to copy, and the offset in the header where the x (image width) and y (image length) are to be placed. `Offsets' are the positions in the image file where various information will be stored. Setting the offsets to -1 will cause them not to be stored. `Bytes to skip' must be a multiple of the no.of bytes per pixel. See manual for details. 25 Help - Multiple click box For each variable (e.g.,red, green, and blue), the minimum and maximum acceptable values are shown. Click on the arrows on each side of the indicator bar to increase or decrease the value, or click and drag the vertical position indicator until the desired value is obtained. 26 Help - Calibrate Select "New calibration" to calibrate an image or background. Select "Hide calibration" to temporarily turn off the calibration window. Select "Un-hide calibration" to turn on the previous calibration window. 27 Help - Grayscale adjustment Changes the mapping of grayscale values to the screen. The grayscale values are also filtered through the currently-selected palette, which allows pseudo-color. This option only has an effect on monochrome and pseudo-color images. It is mainly useful for displaying monochrome images greater than 8 bits per pixel, and for enhancing contrast of low-contrast images. Image white value = the highest pixel value to view in the image. Image black value = the lowest pixel value to view in the image. Screen white level = the highest screen gray level to produce. Screen black level = the lowest screen gray level to produce. 28 Help - Rotate palette Click on "-1" to rotate palette downwards. Click on "1" to rotate palette upwards. 29 Help - Contrast If the image or selected region contains monochrome pixels, select the contrast factor for the monochrome portion of the image. If the image or selected region contains color pixels, select contrast factors for the red, green, and blue components of the color portion. A contrast factor of 100 = no change. Intensity and contrast can also be adjusted interactively by dragging on the palette. 30 Help - Macros Enter one command per line. Any menu item can be part of a macro. In addition, several other non-menu commands can also be used. Many of these commands require string or numeric parameters. The number is usually the same as the order in which the item would appear in the menu or dialog box. See "Macro Programming Guide" in manual for details. Press Shift+Enter to execute a single line. To execute more than one line, click "Execute". The macro will be executed starting at the beginning of the line at which the cursor is positioned. 31 Help - Create image Creates an image by clicking and dragging with the mouse, setting a fixed size and position, or by copying or resizing another image. Use mouse - Use the mouse to select the region for the new image. Fixed size - Enter the x and y size (in pixels), and the x and y position for the upper left corner of the image. Duplicate image - makes an exact copy of current image. Panel->multiframe - Takes an image consisting of several sub-images, and puts each subimage into a multi-frame (3D) image. If 'Manually find frames' is checked, you must click on the upper left corner of each subimage. Otherwise, the subimages are found automatically, using the specified x and y spacing between subimages and the upper left x and y for the first subimage. Multiframe->panel - Converts a multi-frame (3D) image into a large panel consisting of subimages. The spacing (in pixels) between each subimage, and the upper left coordinates for the first subimage in the panel, can be specified. Combine images->multiframe - Similar to panel->multiframe, except each subimage is taken from a different image. These images must all be the same size and bits/pixel. Split multiframe->images - Similar to multiframe->panel, except each frame is put into a separate image. If 'Align images' is checked, the new images are lined up in the specified number of columns. Add frame to image - Adds an additional frame to current image. Use with caution, since not all image file formats support multiple frames. The program will warn you if you try to save a multiframe image in an incompatible format. Grab from screen - Program waits until you click on some other window on the display, then makes a copy of it. Composite 2 or more images - Creates a large single-frame panel image from any number of smaller images. A spreadsheet will appear permitting the position and spacing of each image to be specified. All the images must have the same bits/pixel but can be any size. Resize image window only - changes size of image buffer for currently- selected image, keeping the original image at the upper left corner, and adding black pixels if necessary. This is useful in chopping images to a specific size. 32 Help - Trace curve Click on the image at the leftmost edge of the curve to trace. Curve tracing will proceed from left to right. 33 Help - Measure To measure distance - click at starting point and drag to ending point. To measure angles - click and drag twice to define 2 lines. The angle between the lines in degrees is calculated. (The lines are assumed to extend indefinitely in either direction). 34 Help on Help Click on "Help" to get help. The 'Dismiss' key is inoperative when program is getting input. For more information, please consult the manual. 35 Palette settings No.of colors to leave untouched - Increase this number if flashing occurs when moving the mouse. Decrease the number if colors are displayed incorrectly. 36 Font Click on the appropriate buttons to change the parameters as desired. To select a font not listed, click "other" and type the font specification (for example, *times-medium-r*240*). See your X manual for details. type "xlsfonts" for a list of fonts on your system. 37 Help - 3D Image Specify the total number of frames and the width and height of each frame in pixel units. 38 Help - Sample Palette Click left mouse button on desired color to change foreground color. Click right mouse button on desired color to change background color. Drag the '=' at right to change brightness by moving palette up/down. Drag the two '<' at right to change contrast by squeezing or expanding the palette. Depending on what other applications were running when TN-Image was started, there may be a variable number of fixed colors which belong to other programs and cannot be changed. 39 Help - Scanner Interface Click on "Preview Scan" and set the desired bits/pixsl and the approximate contrast and brightness settings. A low-resolution scan will be performed. Select the desired region and click "Image scan" to scan the image at full resolution. 40 Help - Paste threshold (DOS Version) Select a pixel value to determine which pixels to paste. Pixels above this number will be treated as transparent and will not be pasted. Set the number to its highest value to paste everything. 41 Help - Paste (Unix Version) If no area is selected, the current image will be pasted. Background - > transparent part = Pastes the background pixels into the selected region or image, only in areas where the image is transparent. Background -> opaque part - Pastes the background pixels into the selected region or image, only in areas where the image is opaque. Note that this will make the image disappear. Opaque part -> background - Pastes the opaque pixels in the selected region or image, onto whatever is behind the image. This is the most common use. Transparent part-> background - Pastes the transparent pixels (i.e., those which are not visible) in the selected region or image, onto whatever is behind the image. Stencil -> background - Pastes the transparent pixels in the selected region or image, onto whatever is behind the image, but sets them to the current drawing color first, using the image as a stencil mask. Entire region -> background - Pastes the entire selected region or image onto the background. 42 Help - Copy/Move Click "Copy" to copy a region to another location. Click "Move" to also delete the original. Then select the region to copy by clicking and dragging with the mouse. When the mouse button is released, the selected area will be pasted to the new location. Transparent copy will only paste pixels with values between "Min opaque pixel" and "Max opaque pixel". For indexed-color images, it may be necessary to sort the colormap first. If "Remap pixels" is checked, the copied pixels will be changed to the closest color found in the destination image. The colormap of the destination image is not changed by any copy. Click "Cancel" to stop copying and return to normal mode. 43 Help - Transparency Set the degree of transparency (0=completely opaque, 255=completely transparent) for the currently-selected image. 44 Help - Spray Fine spray sets all the pixels in a fixed area around the cursor to the foreground (drawing) color. Diffuse spray sets the pixels around the cursor to the foreground (drawing) color, but in a diffuse pattern. Math spray allows you to enter any mathematical formula to be applied to all the pixels in a fixed area around the cursor. This can be used to interactively copy small parts of another image or a different part of the image. 'Spray factor' is the size of the fixed area that will be affected. If 'spray only once' is checked, no pixel will be sprayed more than once. Spraying begins when the left mouse button is clicked. To return to normal mode, click the 'cancel' button at left. 45 Help - Rotate Select the degrees to rotate the image or selected area. Antialiasing gives a smoother, more accurate rotation but takes longer. When positioning labels, the foreground and background colors must be different, otherwise the label will be invisible. 46 Help - Spreadsheet Use the gray arrow keys to select the desired pixel, then type the intensity or RGB value. 47 Help - Change image depth Changes the number of bits/pixel in an image. If the bits/pixel of your image is not a multiple of 8, select "maximize contrast" before changing image depths, otherwise the image may become extremely dark. 48 Help - Image attributes Changes the properties of an image. Some background properties, such as colormap, can also be changed by setting "Image Number" to 0. 49 Help - Camera Camera name = Select camera driver to use. See manual for details. Camera driver may require root permissions. Acquisition mode: Normal = Each frame replaces previous frame. Accumulate = Successive frames are averaged. Add = RGB values of successive frames are added (for low-light conditions). Subtract = RGB values of successive frames are subtracted (for removing dark current). Camera Initialize = Resets the camera (may take several seconds) Normal = Ordinary operation Frames Continuous = Acquire video at specified frames/second Single frame = Acquire one frame when START button is clicked. Click Ok or Cancel when finished, START button to begin acquiring images. Click on Single frame to temporarily pause the camera. Post processing command Enter a command or macro to be executed after each frame. Examples: color 10 -20 5 0 = Increase intensity by 10, decrease red by 20, increase green by 5, no change to blue. filter 1 1 10 = High-pass filter, 3x3 kernel, amount of filtering is 10. r+=20; = Image algebra equation 50 Help - Grain/Pattern counting Select region in which to count before starting, or click on the desired image to count all grains or patterns in that image. Clicking on "Count std. size grains" uses the default pattern which is a 3x3 black grain. For all other grains and patterns, click on "Select pattern" and select a typical instance of the pattern, or click on "Read pattern from disk" to use a previously-learned pattern. A higher threshold means pattern-matching is more selective. Making the Mismatch Weight more negative increases the penalty for a mismatched pixel, and will also improve selectivity. For some images, clicking on "Enhance grains" will greatly reduce the error rate. Type of pattern Threshold Match Wt. Mismatch Wt. --------------- --------- --------- ------------ Grains (monochrome) 0.50 1.0 -0.05 Grains (color) 0.45 1.0 -0.10 Grains (enhanced) 0.55 1.0 -0.15 Faces 0.60 1.0 -0.15 Faces (color) 0.95 1.0 -1.00 Patterns 0.90 1.0 -0.80 51 Help - Local contrast maximization This function maximizes the contrast over small areas of the image. Dark areas of the image are lightened and light areas are made darker. The scale setting is the maximum length in pixels of the environment over which the contrast will be maximized. A small value maximizes contrast over a more local area, while a large value gives a more gentle contrast enhancement (and also takes slightly longer). 52 Help - Wavelets Select a wavelet, then click "Decompose" or "Reconstitute" to perform forward or reverse wavelet transform. Changing the reconstitution coefficients causes a smaller range of coefficients to be used in the reverse transform. When filtering low-resolution wavelet components, it is sometimes necessary to add an gray value offset to the image to prevent negative pixel values. Use "Intensity Contrast..Maximize Int" to restore the original contrast levels. 53 Help - Morphological filtering Filters the image or selected area morphologically, by selecting pixels based on their relationship to shapes in the image instead of their brightness. Threshold - sets pixels below threshold to black, above threshold to white. Erosion - eliminates pixels at edge of shapes, making them thinner. Dilation - adds pixels around edges of shapes, making them fatter. Skeletonization - Makes features as thin as possible, ideally a line. Quick segmentation - Draws outlines around features. Watershed segmentation - Finds demarcation points around features. Structuring element - the set of pixels surrounding a given pixel that are checked to determine what to do to the pixel. Type - binary filtering converts image to black + white. Grayscale tries to preserve shades of gray. (Segmentation is always binary). Watershed kernel - set of pixels around a given pixel for watershed segmentation. Contour sep - Sets the spacing between lines for contour maps. 54 Help - Filename Select a filename. Click Chdir to change the current directory. 55 Help Cannot perform this operation on an image greater than 16 bits/pixel, as this would require too much memory. 56 Help This operation should be performed only on 24 bit/pixel images. 57 Help - Total bits/pixel The red + green + blue bits/pixel must add up to the value you specified for total bits/pixel. The total bits/pixel is probably specified incorrectly. 58 Help - Map file A map file is needed to specify what colors the pixel values should be mapped to. See manual for details. 59 Help Enter 'r', 'g', or 'b' here. 60 Help Enter the image number for the image. 61 Help Click 'yes' to replace the existing file. 62 Help - Macro The macro requires a parameter. 63 Help - Resize image Enter a positive number. This does not need to be an integer. The dimension will be multiplied by this factor. 64 Help Enter a string here. 65 Help - Warping Drag the grid points on the image to the desired location to remove distortion in the image. 'c' makes a copy of the current warped image. 'g' toggles the grid on/off. 'r' restores the original rectangular grid. Esc (or clicking the main cancel button) when finished. 66 Help - Curve densitometry Performs densitometry on a curved region one pixel thick. 67 Help - Signal 11 The program has caught signal 11 (SIGSEGV). Save all data immediately. Please send a bug report to tjnelson@helix.nih.gov. 68 Help - Create panel Takes a multi-frame image and creates a composite image with each frame aligned in a rectangular grid of the specified number of columns. 69 Help - Save spreadsheet Saves current contents of spreadsheet in ASCII format. 70 Help - Image notes Enter any notes for the image here. 71 Help - Image registration Warps an image's x and y coordinates so corresponding points on two images are at the same location. All pixels on the warped image are aligned with points on the reference image. The two images do not have to be the same size. Two methods are available: 1. Using user-selected landmark points, a list of points representing known locations on both images. This can be obtained by clicking on "obtain data" or by reading a file from disk. Grain counting files produced by tnimage can also be used. 2. Automatic. Does not need landmark points, but is slower. Outline of procedure for registering 2 images 1. Click 'Image no.' to select the reference and unknown image numbers. 2. Click 'Obtain landmarks' for reference image and select between 1 and 10,000 landmark points. 3. Repeat for Unknown image. 4. If the correspondence between the landmarks on the reference and unknown image is unknown, click "Correlate points". The program will re-sort the landmarks to the best fit. This will be slow if there are over 100 landmarks. 5. If you have a data table of coordinates that needs to be warped or rotated, click "Read match table". See manual for instructions on creating this file. 6. If you want to correct for rotation distortion, click "Calculate rotation map", then "Rotate image". 7. If you want to elastically warp the image, click "Calculate warp vector map", then "Warp image". 8. Alternatively, if you only want to remap the list of coordinates from step 5, without warping the image, click "Create unwarped spot list" or "Create warped spot list" to remap the coordinates to the new alignments. In other words, you can warp an image, or a set of data points, or both. Options Obtain landmks - Interactively define new landmark points on each image. Transl.landmks - Add an x or y value to each landmark Read points - Obtain data points to be warped from disk file Save points - Save data points to disk file Restore points - Restore original data points from backup Read landmarks - Obtain landmark points from disk file Edit landmarks - Modify landmark points (landmarks) Save landmarks - Write landmark points to a disk file Show data vectors - Draws arrows on image showing distance each point will be warped. Correlate points - Permute `unknown' landmark points to ensure the landmark points correspond to each other. This is useful if the landmark points were found using grain counting. Not needed if points were obtained interactively. Calculate rotation map - Create an array so that data points or images can be warped. Must be done before rotating can occur. If a warping is performed, you must recalculate the rotation map. Calculate vector map - Create an array so that data points or images can be warped. Must be done before warping can occur. If a rotation is performed, you must recalculate the vector map. Rotate image - Begin rotation of the warped image. Warp image - Begin registration of the warped image. Smoothing - A higher value creates a smoother rotation map or warp vector map (1=no smoothing). 72 Help - MultiToggleBox Click on desired options. 73 Help - Gradient Removal Red, green, and blue values will be added or multiplied to the image or selected area. The value for each pixel will be calculated by interpolating the values specified for each corner. For multiplications, a value of 100 is no change. 8-bit images (such as GIFs) must be converted to 24 bits/pixel first. 74 Help - Graph list This list shows the areas that you selected by dragging the mouse on the graph. 75 Help - Freetype labels Enter the desired text for the label and click the radio buttons to set the font face, weight, slant, and size. Then click 'Accept' and move to the desired location and click the left mouse button to place the label.