plotimage.pro 30.6 KB
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800
;+
; NAME:
;   PLOTIMAGE
;
; AUTHOR:
;   Craig B. Markwardt, NASA/GSFC Code 662, Greenbelt, MD 20770
;   craigm@lheamail.gsfc.nasa.gov
;
; PURPOSE:
;   Displays an image via a "PLOT"-like interface.
;
; CALLING SEQUENCE:
;   PLOTIMAGE, img, [xrange=xrange,] [yrange=yrange,] ...
;
; DESCRIPTION:
;
;   PLOTIMAGE displays an image (or slice of an image) on the current
;   graphics device.  The syntax is very similar to the PLOT command,
;   in the sense that an XRANGE and YRANGE for the plot can be
;   specified.  
;
;   PLOTIMAGE keeps separate the notions of the image coordinate
;   system and the displayed coordinate system, which allows any input
;   image to be "cropped," "zoomed," or "flipped."
;
;   PLOTIMAGE allows the user to express image extents in physical
;   units rather than pixel units.
;
;   The image coordinate system specifies the physical coordinates of
;   original image data, IMG.  The image is considered to be a 2D
;   array (IMG = ARRAY(NX,NY)), where the values are attached to the
;   midpoint of each geometric pixel.  The image has NX columns and NY
;   rows.  Physical coordinates are attached to each pixel by using
;   the IMGXRANGE and IMGYRANGE keywords.  The IMGXRANGE keyword is a
;   two-element array specifying the "left" and "right" boundaries of
;   the image pixels in physical units; the IMGYRANGE keyword
;   specifies the "top" and "bottom" boundaries of the image.  This is
;   illustrated in Figure 1 for a simplified case.
;
;                                   ___
;         +-----------+-----------+  ^  IMGYRANGE[1]
;         |           |           |  |
;         | IMG[0,1]  | IMG[1,1]  |  |
;         |     +     |     +     |  |
;         |           |           |  |
;         |           |           |  |
;         +-----------+-----------+  |
;         |           |           |  |
;         | IMG[0,0]  | IMG[1,0]  |  |
;         |     +     |     +     |  |
;         |           |           |  |
;         |           |           |  v
;         +-----------+-----------+ ___ IMGYRANGE[0]
;        |                         |
;        |<----------------------->|
;        IMGXRANGE[0]   IMGXRANGE[1]
;
;      Figure 1.  Simplified example of a 2x2 input image,
;      demonstrating that IMG[*,*] values refer to the pixel
;      mid-points, and that IMGXRANGE and IMGYRANGE ranges specify the
;      physical coordinates of the outer edges of the image extent in
;      X and Y, respectively.
; 
;
;   The displayed plot coordinate system is entirely independent of
;   the native image coordinates.  Users can set up the plot scale
;   using any combination of {X,Y}RANGE, {X,Y}STYLE and/or {X,Y}LOG,
;   as they would for any IDL plot, using physical units.  The input
;   image will then be overlayed on this coordinate system.
;
;   If the displayed plot coordinates are narrower than the native
;   image coordinates, then the displayed portion of the image will be
;   cropped to fit.  If the displayed coordinates are wider than the
;   native image coordinates, then the image will be displayed with
;   blank spaces on either side (see Figure 2).  A mirror "flip" is
;   also possible in X and/or Y, if XRANGE or YRANGE are specified in
;   reverse order.
;                                                 ___
;      +---------------------------------------+   ^
;      |            ___                        |   |
;      |             ^  +---------------+      |   |
;      |             |  |               |      |   |
;      |             |  |               |      |   |
;      |    IMGYRANGE|  |     IMG       |      |   | YRANGE
;      |             |  |               |      |   |
;      |             v  |               |      |   |
;      |            ___ +---------------+      |   |
;      |               |<-- IMGXRANGE -->|     |   |
;      |                                       |   v
;      +---------------------------------------+  ___
;     |<-------------   XRANGE   -------------->|
;
;     Figure 2.  Example of an image whose native image coordinates
;     are embedded in a wider plot display range.
;
;   The standard [XY]STYLE keywords can be used to style either axis.
;   However at the very least [XY]STYLE=1 is always implied, i.e. the
;   plot limits exactly obey the [XY]RANGE keyword values.
;
;   If XLOG or YLOG are set, then the image is assumed to be sampled
;   on a logarithmic grid, and logarithmic axes are displayed
;   accordingly.  PLOTIMAGE does not attempt to resample the image
;   from linear scale to logarithmic scale, or reverse.
;
;   Psuedocolor images may be of any type, but must rescaled to a byte
;   range by using the RANGE keyword.  By default the color range used
;   in the rescaling operation is 0 to !D.N_COLORS - 3B.  The extra
;   two color values are reserved for the background and default pen
;   colors.  This behavior can be adjusted by specifying the BOTTOM
;   and/or NCOLORS keywords.
;
;   Truecolor images must always be of type BYTE and one of their
;   dimensions must have 3 elements, corresponding to the three color
;   planes of the image.
;
;
; INPUTS:
;
;   IMG - Array to be displayed.  For single-plane images (i.e.,
;         pseudocolor), the image must be two dimensional and of any
;         real numeric type.  For images that are not of BYTE type,
;         the RANGE keyword must be supplied, and then PLOTIMAGE will
;         rescale the image values to a byte range.
;
;         An image declared as ARRAY(NX,NY) will be NX pixels in the
;         x-direction and NY pixels in the y-direction.  The image is
;         resampled to fill the desired display region (and optionally
;         smoothed).
;
;         For three-plane images (i.e., truecolor) the image must be
;         of type BYTE.  One of the dimensions of the array must have
;         three elements.  Hence it must be one of BYTARR(NX, NY, 3),
;         BYTARR(NX, 3, NY) or BYTARR(3, NX, NY).  The 3-element
;         dimension is recognized automatically.
;
; OPTIONAL INPUTS:
;   NONE
;
; INPUT KEYWORD PARAMETERS:
;
;   IMGXRANGE, IMGYRANGE - Each is a two component vector that
;                          describes the X and Y position of the outer
;                          edges of the first and last pixels.
;                          Default: IMGXRANGE = [0,NX]
;                                   IMGYRANGE = [0,NY]
;
;   XRANGE, YRANGE - Each is a two component vector that specifies the
;                    X and Y plot ranges, respectively.  These values
;                    are not required to coincide with IMG[XY]RANGE.
;                    Default: XRANGE=IMGXRANGE
;                             YRANGE=IMGYRANGE
;
;   POSITION - Position of the inner plot window in the standard
;              graphics keyword format.  Overrides PANEL and SUBPANEL.
;
;   INTERP - if set, interpolate (smooth) the image before displaying.
;            This keyword applies to the screen displays.  For printed
;            images that are coarser than MIN_DPI, the image is
;            implicitly interpolated regardless of INTERP.
;
;   PRESERVE_ASPECT - if set, preserve the aspect ratio of the
;                     original image (in pixels).  The result will be
;                     the largest image that fits in the display
;                     region while maintaining approximately square
;                     pixels.  However, PIXEL_ASPECT_RATIO overrides
;                     PRESERVE_ASPECT.  The POSITION keyword will be
;                     reset upon output to the ultimate image
;                     position.
;                     DEFAULT: not set (image will fill POSITION rectangle)
;
;   PIXEL_ASPECT_RATIO - The ratio of width to height for each pixel.
;                        If specified, then the image will be scaled
;                        so that each pixel has the specified aspect
;                        ratio.  If not specified, then the image will
;                        be scaled independently in X and Y in order
;                        to fill the POSITION rectangle.  NOTE: If you
;                        want to change the overall image aspect
;                        ratio, then use the POSITION keyword.
;                  DEFAULT: undefined (image will fill POSITION rectangle)
;
;   MIN_DPI - if printing, the minimum dot-per-inch pixel resolution
;             for the resulting image.  Output images that would be
;             coarser than this value are resampled to have a
;             resolution of at least MIN_DPI, and smoothed.  Some
;             common resolutions are: screen, 90 dpi; dot matrix, 72
;             dpi; laser printer 300-600 dpi.  Note that large values
;             of MIN_DPI will produce very large output files.
;             Default: 0 (i.e., the output image will not be smoothed)
;
;   RANGE - a two element vector.  If the image is single plane (i.e.,
;           pseudocolor) the input image can be of any real numeric
;           type, and then must be rescaled into byte range with this
;           keyword.  In contrast, truecolor images must always be of
;           type BYTE.  Values are scaled into byte range with the
;           following statement:
;              RESULT = BYTSCL(INPUT, MIN=RANGE(0), MAX=RANGE(1), $
;                              TOP=NCOLORS-1) + BOTTOM
;           so that pixels with an intensity RANGE(0) are set to
;           BOTTOM; those with RANGE(1) are set to the maximum color.
;           Default: no range scaling occurs (and the image must hence
;                    be of type BYTE -- otherwise an error occurs)
;
;   NCOLORS - number of color table values be used in the byte
;             rescaling operation.
;             Default: !D.N_COLORS - BOTTOM - 1 (for default pen color)
;
;   BOTTOM - bottom-most value of the color table to be used in the
;            byte rescaling operation.
;            Default: 1 (for default background color)
;
;   NOERASE - If set, the display is not erased before graphics
;             operations.
;
;   NODATA - If set, the image is not actually displayed, but
;            coordinate axes may be drawn.
;
;   NOAXES - An attempt is made to render the image without coordinate
;            axes.  However, it's usually more straightforward to set 
;            XSTYLE=4 or YSTYLE=4, which is the standard IDL way to
;            disable coordinate axes.
;
;   ORDER - same interpretation as the !ORDER system variable; 
;           if ORDER=0, then the first pixel is drawn in the lower
;           left corner; if ORDER=1, then the first pixel is drawn in
;           the upper left corner.
;           Default: 0
;
;
;   PANEL, SUBPANEL - An alternate way to more precisely specify the
;                     plot and annotation positions.  See SUBCELL.
;
;   PLOTIMAGE will pass other keywords directly to the PLOT command
;   used for generating the plot axes.  XSTYLE=1 and YSTYLE=1 are
;   enforced.
;
; OUTPUTS:
;   NONE
;
; PROCEDURE:
;
; EXAMPLE:
;
;   This example constructs an image whose values are found by
;       z(x,y) = cos(x) * sin(y)
;   and x and y are in the range [-2,2] and [4,8], respectively.
;   The image is then plotted, with appropriate axes.
;
;   x = findgen(20)/5. - 2. + .1   ; 0.1 = half-pixel
;   y = findgen(20)/5. + 4. + .1
;   zz = cos(x) # sin(y)
;   imgxrange = [-2.,2.]           ; extend to pixel edges
;   imgyrange = [4.,8.]
;   plotimage, bytscl(zz), imgxrange=imgxrange, imgyrange=imgyrange
;
;   This second example plots the same image, but with a plot range
;   much larger than the image's.
;
;   xr=[-10.,10]
;   yr=[-10.,10]
;   plotimage, bytscl(zz), imgxrange=imgxrange, imgyrange=imgyrange, $
;      xrange=xr, yrange=yr
;
; SEE ALSO:
;
;   OPLOTIMAGE, BYTSCL
;
; EXTERNAL SUBROUTINES:
;
;   SUBCELL, DEFSUBCELL
;
; MODIFICATION HISTORY:
;   Written, CM, 1997
;   Correct various one-off problems, 02 Feb 1999, CM
;   Made self-contained with some pre-processing, 17 Oct 1999, CM
;   Corrected bug in newly introduced CONGRID functions, 18 Oct 1999, CM
;   Correct behavior with no POSITION keyword, 17 Nov 1999, CM
;   Simplified axis plotting, 17 Nov 1999, CM
;   Use _EXTRA keyword in first PLOT, but with blank TITLEs, 11 Jan
;     2000, CM
;   Correct implementation of X/YSTYLE in first PLOT, 11 Feb 2000, CM
;   Correct CONGRID implementation (small effect when enlarging most
;     images), 14 Feb 2000, CM
;   Major changes: 19 Apr 2000
;      - now handle decomposed color, automatic color mapping via
;        RANGE, and 24-bit multiplane images
;      - new PRESERVE_ASPECT keyword to keep square pixels
;      - removed legacy TVIMAGE code
;      - smoothing is more configurable, esp. for printers, but is not
;        done by default; more printers are supported
;   Corrected INTERPOLATE behavior (thanks to Liam Gumley
;     <Liam.Gumley@ssec.wisc.edu>), other minor tweaks, CM 20 Apr 2000
;   Added ability to use PRESERVE_ASPECT with POSITION, PANEL or
;     SUBPANEL keywords CM 20 Oct 2000
;   Oops, a typo is now fixed, CM 23 Oct 2000
;   Add fix for MacIntoshes and DECOMPOSED color, Tupper, 02 Aug 2001
;   Better behavior with fractional pixels (ie, when the image pixels
;     are very large compared to the screen), 23 Aug 2001
;   Add support for Z buffer, CM, 20 Oct 2002
;   Memory conservation: use REVERSE() to reverse IMG; rewrote
;     PLOTIMAGE_RESAMP to rescale entire image instead of each color plane
;     separately.  Jeff Guerber, 30 July 2003
;   Add PIXEL_ASPECT_RATIO keyword, 22-25 Nov 2005
;   Check for the case of an 1xNXxNY 3D image and treat it as a 2D
;     image.  The "1" dimension can be anywhere, CM, 03 Sep 2006
;   Add the ORDER keyword parameter, CM, 20 Mar 2007
;   Enable XLOG and YLOG keywords, for logarithmic axes;
;     doesn't actually resample the image from linear<->log, CM
;     21 Jan 2009
;   Documentation, CM, 21 Jan 2009
;   Allow reverse color scale, CM, 13 Nov 2010
;   Add checking for decomposed state of Z buffer, thanks David
;     Palmer, 2013-08-12
;
;   $Id: plotimage.pro,v 1.16 2013/09/30 02:22:49 cmarkwar Exp $
;
;-
; Copyright (C) 1997-2001,2003,2005,2006,2007,2009,2010,2013 Craig Markwardt
; This software is provided as is without any warranty whatsoever.
; Permission to use, copy, modify, and distribute modified or
; unmodified copies is granted, provided this copyright and disclaimer
; are included unchanged.
;-
;%insert HERE
;%include subcell.pro
;%include defsubcell.pro

; Utility routine to resample an image
;
;  IMAGE - image data ARRAY(NX,NY,BDEPTH)
;  NX,NY - original X,Y image size
;  BDEPTH- original image depth
;  NEWX, NEWY- desired X,Y image size
;  INTERP - if set, then use bilinear interpolation, otherwise nearest neighbor
function plotimage_resamp, image, nx, ny, bdepth, newx, newy, interp=interp

  ;; Sometimes the final dimension is lost.  Put it back
  image = reform(image, nx, ny, bdepth, /overwrite)

  ;; Correct interpolation
  srx = float(nx)/newx * findgen(newx) - 0.5 + 0.5*(float(nx)/newx)
  sry = float(ny)/newy * findgen(newy) - 0.5 + 0.5*(float(ny)/newy)
  srz = indgen(bdepth)
  if keyword_set(interp) then $
    return, interpolate(image, srx, sry, srz, /grid)

  ;; Simple nearest neighbor interpolation
  return, interpolate(image, round(srx), round(sry), srz, /grid)
end

pro plotimage_pos, xrange0, imgxrange0, imgxsize, xreverse, srcxpix, imgxpanel, $
                   logscale=logscale, $
                   quiet=quiet, status=status, pixtolerance=pixtolerance

  if keyword_set(logscale) then begin
     if min(xrange0) LE 0 OR min(imgxrange0) LE 0 then $
        message, ('ERROR: if XLOG or YLOG is set, then the image boundary cannot '+$
                  'cross or touch zero.  Did you forget to set IMGXRANGE or IMGYRANGE?')
     xrange    = alog10(xrange0)
     imgxrange = alog10(imgxrange0)
  endif else begin
     xrange    = xrange0
     imgxrange = imgxrange0
  endelse

  if n_elements(pixtolerance) EQ 0 then pixtolerance = 1.e-2
  status = 0
  ;; Decide if image must be reversed
  xreverse = 0
  if double(xrange(1)-xrange(0))*(imgxrange(1)-imgxrange(0)) LT 0 then begin
      xreverse = 1
      imgxrange = [imgxrange(1), imgxrange(0)]
  endif

  srcxpix  = [ 0L, imgxsize-1 ]
  ;; Size of one x pix
  dx = double(imgxrange(1) - imgxrange(0)) / imgxsize

  if min(xrange) GE max(imgxrange) OR max(xrange) LE min(imgxrange) then begin
      message, 'WARNING: No image data in specified plot RANGE.', /info, $
        noprint=keyword_set(quiet)
      return
  endif

  ;; Case where xrange cuts off image at left
  if (xrange(0) - imgxrange(0))/dx GT 0 then begin
      offset = double(xrange(0)-imgxrange(0))/dx
      if abs(offset-round(offset)) LT pixtolerance then $
        offset = round(offset)
      srcxpix(0) = floor(offset)
      froffset = offset - floor(offset)
      if abs(froffset) GT pixtolerance then begin
          xrange = double(xrange)
          xrange(0) = imgxrange(0) +dx*srcxpix(0)
      endif
  endif

  ;; Case where xrange cuts off image at right
  if (xrange(1) - imgxrange(1))/dx LT 0 then begin
      offset = double(xrange(1)-imgxrange(0))/dx
      if abs(offset-round(offset)) LT pixtolerance then $
        offset = round(offset)
      srcxpix(1) = ceil(offset) - 1
      froffset = offset - ceil(offset)
      if abs(froffset) GT pixtolerance then begin
          xrange = double(xrange)
          srcxpix(1) = srcxpix(1) < (imgxsize-1)
          xrange(1) = imgxrange(0) + dx*(srcxpix(1)+1)
      endif
  endif

  imgxpanel = [0., 1.]
  if (xrange(0) - imgxrange(0))/dx LT 0 then $
    imgxpanel(0) = (imgxrange(0) - xrange(0))/(xrange(1)-xrange(0))
  if (xrange(1) - imgxrange(1))/dx GT 0 then $
    imgxpanel(1) = (imgxrange(1) - xrange(0))/(xrange(1)-xrange(0))

  status = 1
  return
end

;; Main program
pro plotimage, img0, xrange=xrange0, yrange=yrange0, $
               imgxrange=imgxrange0, imgyrange=imgyrange0, $
               xlog=xlog, ylog=ylog, $
               position=position, panel=panel, subpanel=subpanel, $
               xstyle=xstyle, ystyle=ystyle, title=title, $
               interp=interp0, quiet=quiet, dither=dither, $
               preserve_aspect=paspect, pixel_aspect_ratio=asprat, $
               min_dpi=min_dpi, order=order, $
               ncolors=ncolors0, bottom=bottom0, range=range, $
               noerase=noerase0, nodata=nodata, noaxes=noaxes, $
               pixtolerance=pixtolerance, _EXTRA=extra

  ;; Return to user when an error is encountered
  on_error, 2

  ;; Usage message
  if n_params() EQ 0 then begin
      message, 'PLOTIMAGE, image, xrange=, yrange=, imgxrange=, imgyrange=,..', $
               /info
      return
  endif

  ;; Must have a byte-scaled image already
  imgsize  = size(img0)

  ;; Make sure windowing exists (borrowed from IMDISP)
  if ((!d.flags and 256) ne 0) and (!d.window lt 0) then begin
      window, /free, /pixmap
      wdelete, !d.window
  endif

  ;; Parameter checking
  if n_elements(ystyle) EQ 0 then ystyle = 0L
  if n_elements(xstyle) EQ 0 then xstyle = 0L
  if keyword_set(nodata) then mynodata = 1 else mynodata = 0
  if n_elements(pixtolerance) EQ 0 then pixtolerance = 1.e-2
  if n_elements(title) EQ 0 then title = ''
  if n_elements(min_dpi) EQ 0 then min_dpi = 0
  interp = keyword_set(interp0)
  noerase = keyword_set(noerase0)
  imgpanel = [0., 0., 1., 1.]

  ;; Default handling of color table stuff
  if n_elements(bottom0) EQ 0 then bottom0 = 1B
  bottom = byte(bottom0(0)) < 255B
  dncolors = min([!d.n_colors, !d.table_size, 256])
  if n_elements(ncolors0) EQ 0 then ncolors0 = dncolors - 1 - bottom
  ;; Make sure color table values are in bounds
  ncolors = floor(ncolors0(0)) < 256
  if bottom + ncolors GT 256 then ncolors = 256 - bottom

  ;; Image size and dimensions
  nimgdims  = imgsize(0)
  imgtype   = imgsize(nimgdims+1)
  if nimgdims LT 2 OR nimgdims GT 3 then begin
      message, 'ERROR: image must have 2 or 3 dimensions'
  endif

  if nimgdims EQ 2 then begin
      ;; Two dimensional image is pseudo color
      img = img0

      ONE_CHANNEL_IMAGE:
      imgxsize = imgsize(1)
      imgysize = imgsize(2)
      bdepth = 1

      if imgtype NE 1 then begin
          if n_elements(range) LT 2 then $
            message, 'ERROR: non-byte image must be scaled with RANGE keyword'
          if range(0) LE range(1) then begin
              img = bytscl(img, min=range(0), max=range(1), top=ncolors-1B) $
                + bottom
          endif else begin
              ;; Reverse color scheme
              img = bytscl(img, min=range(1), max=range(0), top=ncolors-1B)
              img = ncolors-1B-img + bottom
          endelse
      endif
      img = reform(img, imgxsize, imgysize, bdepth, /overwrite)
  endif else begin
      wh = where(imgsize(1:3) EQ 1, ct)
      if ct GT 0 then begin
          imgxsize = 1
          imgysize = 1
          j = 0
          for i = 1, 3 do if imgsize(i) NE 1 then begin
              if j EQ 0 then imgxsize = imgsize(i) else imgysize = imgsize(i)
              j = j + 1
          endif
          img = reform(img0, imgxsize, imgysize)
          imgsize = size(img)

          goto, ONE_CHANNEL_IMAGE
      endif else begin
          ;; Three dimensional image has three planes
          wh = where(imgsize(1:3) EQ 3, ct)
          if imgtype NE 1 then $
            message, 'ERROR: true color image must of type byte'
          if ct EQ 0 then $
            message, ('ERROR: True color image must have 3 elements '+$
                      'in one of its dimensions')
          truedim = wh(0)
          
          ;; Shuffle the data so planes are interleaved ...
          case truedim of
              0: img = transpose(img0, [1,2,0]) ;; ... from pixels interleaved
              1: img = transpose(img0, [0,2,1]) ;; ... from rows interleaved
              2: img = img0                 ;; ... by straight copying
          end

          imgsize = size(img)
          imgxsize = imgsize(1)
          imgysize = imgsize(2)
          bdepth = imgsize(3)

      endelse

  endelse

  ;; By default, we have no info about the image, and display the
  ;; whole thing
  if n_elements(imgxrange0) LT 2 then imgxrange = [ 0., imgxsize ] $
  else imgxrange = 0. + imgxrange0(0:1)
  if n_elements(xrange0) LT 2 then xrange = imgxrange $
  else xrange = 0. + xrange0(0:1)

  status = 0
  plotimage_pos, xrange, imgxrange, imgxsize, xreverse, srcxpix, imgxpanel, $
    quiet=keyword_set(quiet), status=status, pixtolerance=pixtolerance, $
    logscale=xlog
  if status EQ 0 then mynodata = 1 $
  else imgpanel([0,2]) = imgxpanel

  ;; By default, we have no info about the image, and display the
  ;; whole thing
  if n_elements(imgyrange0) LT 2 then imgyrange = [ 0., imgysize ] $
  else imgyrange = 0. + imgyrange0(0:1)
  if n_elements(yrange0) LT 2 then yrange = imgyrange $
  else yrange = 0. + yrange0(0:1)
  if keyword_set(order) then yrange = [yrange(1), yrange(0)]

  status = 0
  plotimage_pos, yrange, imgyrange, imgysize, yreverse, srcypix, imgypanel, $
    quiet=keyword_set(quiet), status=status, pixtolerance=pixtolerance, $
    logscale=ylog
  if status EQ 0 then mynodata = 1 $
  else imgpanel([1,3]) = imgypanel

  ;; Dimensions of output image in pixels
  nx = srcxpix(1)-srcxpix(0)+1
  ny = srcypix(1)-srcypix(0)+1

  ;; Create a coordinate system by plotting with no data or axes
  if n_elements(position) EQ 0 AND n_elements(panel) EQ 0 AND $
    n_elements(subpanel) EQ 0 then begin

      ;; If PANEL/SUBPANEL is not given, then plot once to set up
      ;; axes, despite NOAXES
      plot, xrange, yrange, noerase=noerase, /nodata, $
        xstyle=xstyle OR 5, ystyle=xstyle OR 5, xlog=xlog, ylog=ylog, $
        xrange=xrange, yrange=yrange, xtitle='', ytitle='', title='', $
        _EXTRA=extra

      ;; Retrieve axis settings
      xwindow = !x.window
      ywindow = !y.window

      subpanel1 = [xwindow(0), ywindow(0), xwindow(1), ywindow(1)]
      imgposition = subcell(imgpanel, subpanel1)
      position = subpanel1

  endif else begin

      ;; Construct the plot size from panel info.  Default is full-screen
      if NOT keyword_set(noerase) then erase
      if n_elements(position) GE 4 then begin
          imgposition = subcell(imgpanel, position)
      endif else begin
          if n_elements(panel) LT 4 then panel = [0.0,0.0,1.0,1.0]
          if n_elements(subpanel) LT 4 then subpanel = [-1., -1, -1, -1]
          subpanel = defsubcell(subpanel)

          imgposition = subcell(subcell(imgpanel, subpanel), panel)
          position = subcell(subpanel, panel)
      endelse

      xwindow = position([0,2])
      ywindow = position([1,3])

  endelse

  ;; If the aspect is to be preserved then we need to recompute the
  ;; position after considering the image size.  Since we have already
  ;; computed the outer envelope of the image from either the POSITION
  ;; or PANEL, or from the plot window itself, we can now go to the
  ;; logic which estimates the aspect-corrected size.

  if (keyword_set(paspect) OR n_elements(asprat) GT 0) AND $
    nx GT 0 AND ny GT 0 then begin

      if n_elements(asprat) EQ 0 then asprat1 = 1.0 $
      else asprat1 = asprat(0) + 0.

      ;; If we are preserving the aspect, then re-plot after scaling
      ;; the POSITION

      imgaspect = float(ny)/float(nx)/asprat1
      dispaspect = (ywindow(1)-ywindow(0))*!d.y_vsize $
        / ((xwindow(1)-xwindow(0))*!d.x_vsize)

      ;; Compute the new image dimensions
      if imgaspect GT dispaspect then begin
          x0 = total(xwindow)/2
          dx = (ywindow(1)-ywindow(0))*!d.y_vsize/(imgaspect*!d.x_vsize)
          xwindow = x0 + dx*[-0.5,0.5]
      endif else begin
          y0 = total(ywindow)/2
          dy = (xwindow(1)-xwindow(0))*!d.x_vsize*imgaspect/!d.y_vsize
          ywindow = y0 + dy*[-0.5,0.5]
      endelse

      subpanel1 = [xwindow(0), ywindow(0), xwindow(1), ywindow(1)]
      imgposition = subcell(imgpanel, subpanel1)
      position = subpanel1

      ;; Replot to regain coordinate system
      plot, xrange, yrange, /noerase, /nodata, $
        xstyle=xstyle OR 5, ystyle=xstyle OR 5, xlog=xlog, ylog=ylog, $
        xrange=xrange, yrange=yrange, xtitle='', ytitle='', title='', $
        position=position, _EXTRA=extra

  endif

  ;; Draw the image data
  if NOT keyword_set(mynodata) then begin

      ;; Reverse X- or Y- directions if necessary
      if xreverse then $
        srcxpix = imgxsize - 1 - [srcxpix(1), srcxpix(0)]
      if yreverse then $
        srcypix = imgysize - 1 - [srcypix(1), srcypix(0)]

      ;; Extract relevant image elements
      img = (temporary(img))(srcxpix(0):srcxpix(1), srcypix(0):srcypix(1),*)
      img = reform(img, nx, ny, bdepth, /overwrite)

      ;; Complete the extraction, if reversed
      if xreverse then begin
          img = reverse(img, 1, /overwrite)
          img = reform(img, nx, ny, bdepth, /overwrite)
      endif
      if yreverse then begin
          img = reverse(img, 2, /overwrite)
          img = reform(img, nx, ny, bdepth, /overwrite)
      endif

      ;; Compute the image position on screen in pixels
      x0 = round(imgposition(0) * !d.x_vsize)
      y0 = round(imgposition(1) * !d.y_vsize)
      dx = round((imgposition(2) - imgposition(0)) * !d.x_vsize) > 1
      dy = round((imgposition(3) - imgposition(1)) * !d.y_vsize) > 1

      ;; Decide which output type
      windowing = (!d.name EQ 'WIN') OR (!d.name EQ 'MAC') OR (!d.name EQ 'X')
      printing = (!d.name EQ 'PRINTER') OR (!d.flags AND 1) NE 0

      ;; Decide whether to resample the image
      rescaling = (windowing OR (!d.name EQ 'Z')) $
        AND ((dx NE nx) OR (dy NE ny))

      ;; If printing, and the printed resolution of the image will be
      ;; too coarse, then we should resample and interpolate
      dpi = min([nx*!d.x_px_cm/dx, ny*!d.y_px_cm/dy]*2.54) ; d.p.i. of image
      dxsize = dx & dysize = dy
      if printing AND (dpi LT min_dpi(0)) then begin
          dx = round(min_dpi(0)*dx/(2.54*!d.x_px_cm)) > nx
          dy = round(min_dpi(0)*dy/(2.54*!d.y_px_cm)) > ny
          interp = 1
          rescaling = 1
      endif

      ;; Rescale the image if needed
      if rescaling then begin
          img = plotimage_resamp(temporary(img), nx, ny, bdepth, $
            dx, dy, interp=interp)
          img = reform(img, dx, dy, bdepth, /overwrite)
      endif

      ;; Generic printer device
      if !d.name EQ 'PRINTER' then begin
          if bdepth EQ 3 then begin
              device, /true_color
              tv, img, x0, y0, xsize=dxsize, ysize=dysize, true=3
          endif else begin
              device, /index_color
              tv, img, x0, y0, xsize=dxsize, ysize=dysize
          endelse
          goto, DONE_IMG
      endif

      ;; Devices with scalable pixels
      if (!d.flags AND 1) NE 0 then begin
          if bdepth EQ 3 then begin
              tvlct, r, g, b, /get
              loadct, 0, /silent
              tv, img, x0, y0, xsize=dxsize, ysize=dysize, true=3
              tvlct, r, g, b
          endif else begin
              tv, img, x0, y0, xsize=dxsize, ysize=dysize
          endelse
          goto, DONE_IMG
      endif

      ;; Get visual depth (in bytes) and decomposed state
      decomposed0 = 0
      vdepth = 1
      version = float(!version.release)
      if windowing then begin

          ;; Visual depth
          if version GE 5.1 then begin
              device, get_visual_depth=vdepth
              vdepth = vdepth / 8
          endif else begin
              if !d.n_colors GT 256 then vdepth = 3
          endelse

          ;; Decomposed state
          if vdepth GT 1 then begin
              if version GE 5.2 then device, get_decomposed=decomposed0
              if bdepth EQ 3 then    device, decomposed=1 $
              else                   device, decomposed=0
          endif
      endif else if !d.name EQ 'Z' then begin
          if !d.n_colors GT 256 then vdepth = 3
      endif

      ;; If visual is 8-bit but image is 24-bit, then quantize
      if vdepth LE 1 AND bdepth EQ 3 then begin
          img = color_quan(temporary(img), 3, r, g, b, colors=ncolors-1, $
                           dither=keyword_set(dither)) + bottom
          tvlct, r, g, b, bottom
          bdepth = 1
      endif

      ;; Put the image
      if bdepth EQ 3 then tv, img, x0, y0, true=3 $
      else                tv, img, x0, y0

      ;; Restore the decomposed state
      if windowing then begin
          if vdepth GT 1 then device, decomposed=decomposed0
          ;; Tupper supplies following work-around for MacIntoshes
          if (!d.name EQ 'MAC') then tv, [0], -1, -1
      endif
  endif

  ;; Plot the axes if requested
  DONE_IMG:
  if NOT keyword_set(noaxes) then begin
      if n_elements(xrange) EQ 0 then begin
          if n_elements(imgxrange) GT 1 then xrange=imgxrange $
          else xrange = [0L, imgxsize]
      endif
      if n_elements(yrange) EQ 0 then begin
          if n_elements(imgyrange) GT 1 then yrange=imgyrange $
          else yrange = [0L, imgysize]
      endif

      plot, xrange, yrange, /noerase, /nodata, /normal, $
        xrange=xrange, yrange=yrange, xlog=xlog, ylog=ylog, $
        xstyle=xstyle OR 1, ystyle=ystyle OR 1, title=title, $
        position=position, _EXTRA=extra
  endif

  return
end