image-processing

Outdated egg!

This is an egg for CHICKEN 4, the unsupported old release. You're almost certainly looking for the CHICKEN 5 version of this egg, if it exists.

If it does not exist, there may be equivalent functionality provided by another egg; have a look at the egg index. Otherwise, please consider porting this egg to the current version of CHICKEN.

This page is maintained in the package's github repository.

This is a Chicken Scheme egg with various image processing utilities.

Note that a few of these functions call ImageMagick.

Color spaces

*max-red*constant
*max-green*constant
*max-blue*constant
*max-grey*constant
*max-hue*constant
*max-saturation*constant
*max-value*constant: These are exposed in case we want to switch to 16-bit grayscale or colors.

rgb->hsv rgbprocedure
hsv->rgb hsvprocedure
rgb->cd rgbprocedure
rgb->cmyk rgbprocedure
rgb->uv-hsv cprocedure: Various color space conversions. Colors are vectors of values.

rgb->xyz cprocedure
xyz->rgb cprocedure
xyz->l*ab cprocedure
l*ab->xyz cprocedure
rgb->l*ab cprocedure
l*ab->rgb cprocedure: Convert to l*ab. Colors are vectors of values. This is space in which Euclidean distance is similar to human perceputal distance. These implement a 2 degree observer and the D65 illuminant.

rgb->html cprocedure: Not really a color space conversion, covert an rgb color to an HTML color #rrggbb.

Images

define-structure pbm raw? bitmaprecord
define-structure pgm raw? maxval greyrecord
define-structure ppm raw? maxval red green bluerecord
pnm? mprocedure: Three image formats are supported. PBMs are binary images, PGMs are greyscale and PPMs are tri-plane (nominally RGB) images. All planes are stored as scheme matrices. Together all of these formats are called PNM. This is a standard format that many tools (such browser and ImageMagick) can read and write. PNMs can be serialized as either raw packed binaries or ascii text files. Note that we are unable to read or write raw PBMs.

image-ref i pprocedure: Index into an PNM image or into a matrix. This is like indexing into a matrix except that the arguments are reversed. A matrix-ref takes the row and then the column, image-ref takes a point which is an (x,y) coordinate.

pbm-ascii pbmprocedure: Ensure that the PBM image is in ascii. We are unable to read or write raw PBMs.

pnm-copy pnmprocedure: The resulting image will not share any data with the original.

Creating images

read-pnm pathnameprocedure: Read a PNM. We determine the kind of image (PBM, PGM, PPM see above) from the file header.

write-pnm pnm pathnameprocedure: Write a PNM. If don't provide an extension an appropriate one will be appended to your pathname.

pbm-constant width height bitprocedure
pgm-constant width height greyprocedure
ppm-constant width height red green blueprocedure: Create constant images.

pbm-left-vertical-stripe width height leftprocedure: Create a black vertical stripe on a white image.

crop-image pnm x y width heightprocedure: Crop an image and return a new image of the given size.

pnm->ppm pnmprocedure
pnm->pgm pnmprocedure: Convert any image format to a color or greyscale image

pbm->pgm pbmprocedure
pgm->ppm pgmprocedure
pbm->ppm pbmprocedure
ppm->pgm ppmprocedure: Various image conversions.

pgm->pbm pgm thresholdprocedure
ppm->pbm ppm thresholdprocedure: Convert a color or greyscale image to a binary image with the provided threshold.

Basic image information

pnm-width pnmprocedure
pnm-height pnmprocedure: Basic image information.

ppm-hue ppmprocedure
ppm-saturation ppmprocedure
ppm-value ppmprocedure: Creates a greyscale image representing one of the HSV channels from a color image.

ppm-mean image colour-transformprocedure
ppm-covariance image colour-transformprocedure
pgm-mean imageprocedure
pgm-variance imageprocedure: Compute the mean or covariance of an image. These take an optional colour transform.

Logical and morphological operations

pbm-and pbm1 pbm2procedure
pbm-or pbm1 pbm2procedure
pbm-not pbmprocedure
pbm-xor pbm1 pbm2procedure: Basic logical operations on binary images.

pgm-absolute-difference pgm1 pgm2procedure: Compute the difference between two greyscale images.

empty-pnm? pnmprocedure: Is this image empty?

pbm-skeletonize pbmprocedure
pbm-bloat pbm nprocedure: Bloat or skeletonize a binary image.

pbm-flood pbm pointprocedure: Floor fill a binary image.

PNM movies

pnm-movie-frame-pathname pathname iprocedure
pnm-movie-length pathnameprocedure
read-pnm-movie pathnameprocedure
write-pnm-movie pnm-movie pathnameprocedure: A PNM movie is a sequence of PNMs on disk.

Image operations

overlay-pbm-on-pnm pbm pnmprocedure: The white pixels in the pbm become white in the resulting image.

pnm-black-window pnm upper-left lower-rightprocedure
pnm-white-window pnm upper-left lower-rightprocedure: Draw white or black windows on an image.

pbm-ppm-and pbm ppmprocedure
pgm-and-pbm pgm pbmprocedure: Black pixels in the PBM become black pixels in the result.

pnm-shift pnm deltaprocedure: Shift an image down and rightward by the give number of pixels.

pgm-smooth pgm sigmaprocedure: Smooth a greyscale image with the given variance.

pbm-proximity-clusterer pbm thresholdprocedure: Cluster white pixels in the binary image with the given distance threshold.

pnm-rotate pnmprocedure
pnm-flip pnmprocedure: Rotate (transpose) an image or flip (mirror) an image.

ppm-burn base mask colourprocedure: Burn the ppm with the given mask and color.

Image pairs

normal-flow-magnitude pgm1 pgm2 epsilon sigma sensitivityprocedure
threshold-normal-flow-magnitude pgm1 pgm2 epsilon sigma thresholdprocedure: Compute normalized flow between two pairs of greyscale images.

Thresholding

flatten-ppm ppm colour-transformprocedure: Returns a single matrix from a tri-plane PPM. This maps colour-transform over the image.

binary-threshold pgm thresholdprocedure: Threshold a greyscale image.

find-threshold-otsu pgmprocedure
binary-threshold-optimal pgmprocedure: Otsu's Method to find an optional threshold and apply it.

find-threshold-means pgmprocedure
binary-threshold-means pgmprocedure: Method of Successive Means to find an optimal threshold and apply it.

Colour Threshold

sample-image ppm colour-tx window-centre window-sizeprocedure: Map colour-tx over an region of a color image.

binary-threshold-colour ppm colour-tx point thresholdprocedure: Binary threshold a colour image by sampling a region around the given point and thresholding all values that are within the given distance threshold of the mean color of that region.

colour-threshold ppm colour-tx mu isigma thresholdprocedure: Binary threshold a colour image with the given mean and variance. The mahalanobis-distance between each pixel and the given mean and variance will be computed and compared to the threshold.

Histogram Equalization

histogram-equalise pgmprocedure: Equalize a greyscale image.

Adaptive Thresholding

make-integral-matrix matrixprocedure: Given a matrix compute an integral matrix which can provide the mean value of a rectangular region in O(1).

compute-integral-matrix-mu integral-matrix x y wprocedure
compute-integral-matrix-sigma squared-integral-matrix mu x y wprocedure: This computes the mean and variance of a square of size w in the given integral matrix.

compute-adaptive-threshold integral-matrix squared-integral-matrix x y wprocedure: Compute an adaptive threshold on the given integral matrix.

adaptive-threshold pgm winsizeprocedure: Compute an adaptive threshold for a greyscale image with a given window size.

slide-window m f sizeprocedure: Map f over a matrix passing it square submatrices of the given size. At the edges f is passed #f.

count-pixels aprocedure: Count number of true values in a matrix.

ppm->label-closest ppm mu1 sigma1 mu2 sigma2 #!optional colour-transformprocedure: Create a binary image from a color image where points are either white or black depending on which is closer to the given means and variances.

Histogram

find-histogram pixmap maxvalprocedure
histogram-mean histogram iprocedure
histogram-variance histogram mu iprocedure
normalised-histogram histogram valprocedure
weighted-histogram histogramprocedure
cumulative-histogram histogramprocedure
find-between-class-variances omegas mus mu-totalprocedure: Compute histograms from a matrix.

Rendering Line Segments

midpoint lprocedure: Compute the midpoint of a line segment.

orientation vprocedure: Compute the orientation of a vector.

line-segment-orientation lprocedure: Compute the orientation of a line segment.

line-segment->points lprocedure
line-segments->points lsprocedure
points->line-segments psprocedure: Convert between points and line semgents.

Points and bounding boxes

points->pbm-of-size points height widthprocedure: Take the list of points and create a binay image of the given size where those points are white.

points-bounding-box pointsprocedure: Computing a bounding box for a set of points.

bounding-box-size bbprocedure: Compute the size of a bounding box

points->points-bounding-box points bbprocedure: Recompute the coordinates of the given points relative to the top-left corner of the bounding box.

points->bounding-box-points pointsprocedure: Convert points to a list of points on the bounding box of those points.

normalize-to-bounding-box psprocedure
normalize-to-other-bounding-box points psprocedure
points->target-bounding-box points target-bbprocedure
points->other-target-bounding-box points ps target-bbprocedure: Normalize the cooredinates of these points to their bounding box. Or to the bounding box of the given points.

bounding-box-bloat bb pprocedure: Bloat a bounding box.

bounding-box-crop bb imageprocedure: Crop a bounding box out of an image.

pbm->points pbmprocedure
points->pbm points height widthprocedure: Convert a binary image to a set of points or vice versa.

Quantization

quantize-coordinate xprocedure
quantize-point pprocedure
quantize-points psprocedure
quantize-line-segment lprocedure
quantize-line-segments lsprocedure: Quantize coordinates, points or line segments.

Ellipses

define-structure ellipse x0 y0 t0 a bprocedure
ellipse-center ellipseprocedure
ellipse-area ellipseprocedure
ellipse-eccentricity ellipseprocedure: Define and compute basic properties of ellipses.

radial-distance theta phiprocedure: The radial distance between two angles.

point-on-ellipse? p ellipse toleranceprocedure: Compute if a point is on an ellipse.

line-segment->ellipse lprocedure: Place an ellipse around a line segment

ellipse->points e #!optional (n 360)procedure: Sample n points on an ellipse.

Resizing

resize-image w h iprocedure: Resize an image. This is quite slow as it writes the image out and calls imagemagick to resize it.

subsample-pbm pbmprocedure: Subsample a pbm by a factor of 2.

scale-ppm ppm scaleprocedure: Scale a pbm by a factor of 2 or 4.

Connected Components

connected-component-filter pbm delta thresholdprocedure: Compute connected components of a binary image with distance thershold delta and size threshold threshold.x

define-private-structure vertex pixels vertex edgesrecord
define-private-structure edge u vrecord
define-private-structure graph vertices edgesrecord: These are defined private for now as they need to be renamed.

pbm->graph pbm deltaprocedure
labeling->graph labeling deltaprocedure: Convert a binary image to a graph where pixels within delta of each other are connected.

dereference-vertex uprocedure
connected-components gprocedure: Compute the connect components of a graph.

vertices->pbm vertices height widthprocedure
graph->pbm g height widthprocedure: Convert a graph or list of vertices to a pbm.

Chains

pbm->chains pbmprocedure: Convert a binary image to a list of chains, pixels that are connected to each other.

chains->pbm chains height widthprocedure: Convert a list of chains to a pbm.

chain-filter pbm thresholdprocedure: Filters chains below a certain length.

break-chain chain lprocedure
break-chains chains lprocedure: Break chains below a certain length.

Conjuring

conjure pbms delta span threshold1 threshold2procedure: Fill in missing values in pairs of frames.

Distance Transform

distance-transform pbmprocedure
closest-transform-ref closest-transform pprocedure: Compute the Manhattan distance transform for a binary image and reference into that distance transform.

euclidean-1d-dt vprocedure
euclidean-2d-dt mprocedure
euclidean-1d-dt-vals vprocedure
euclidean-2d-dt-vals mprocedure: Compute a generalized squared-euclidean distance transform on the given vector or matrix. The -vals version also returns a map of the closest points.

Buffers

define-structure pnm-buffer buffer width height pixfmt storage typerecord: A mechanism to store raw image pointers.

pnm->pixfmt p p4?procedure
pixfmt->stride pixfmtprocedure
image-type->pixfmt t p4?procedure
image-type->storage-size t p4?procedure
pnm->image-type pprocedure
pnm->storage pprocedure
pixfmt->red pixfmtprocedure
pixfmt->green pixfmtprocedure
pixfmt->blue pixfmtprocedure
pnm-buffer-size p p4?procedure
pnm->pnm-buffer! p #!rest p4?procedure
pnm-fill-buffer! p b pixfmtprocedure
free-pnm-buffer! pprocedure: Misc raw image buffer operations.

imlib-image->pnm-buffer! imageprocedure
imlib-image->ppm imageprocedure
ppm->imlib-image ppmprocedure: Convert between PPMs and Imlib images. PGMs and PBMs must first be converted to PPMs because Imlib only handles RGB images.

pnm-buffer->pnm bprocedure
pnm-buffer->pbm pprocedure
pnm-buffer->pgm pprocedure
pnm-buffer->ppm pprocedure: Convert a raw image buffer to a scheme image.

image->pnm-buffer! imageprocedure: Convert any image to a raw image buffer.

Misc

ppm-absolute-difference ppm1 ppm2procedure: Compute the absolute difference between two color images.

set-ppm-pixel! ppm x y valueprocedure: Set a pixel in a color image to a given value (a 3-vector).

pnm-pixel? i x yprocedure: Is this a valid position in the given image?

map-ppm-values ppm fprocedure: Map f on each component of each pixel in the image.

Stacking

pbm-stack-vertical pbm1 pbm2procedure
pbm-stack-horizontal pbm1 pbm2procedure
ppm-stack-vertical ppm1 ppm2procedure: Stack pbms.

ppm-stack-horizontal ppm1 ppm2procedure: Stack ppms.

Angles

degrees->radians angleprocedure
radians->degrees angleprocedure
normalize-rotation rotationprocedure
rotation+ x yprocedure
rotation- x yprocedure
angle-separation x yprocedure
mean-angle anglesprocedure: Misc operation on angles.

rotation-matrix-2d thetaprocedure: Produce the 2D rotation matrix for the given angle.

Display

show iprocedure: Display an image. This invokes feh.

License

  Written by the
  [[Center for the Computational Study of Cognition, Control, and Perception|http://upplysingaoflun.ecn.purdue.edu/~qobi/cccp/]].

  Maintainer: Andrei Barbu, andrei@0xab.com

  Copyright 1993-1995 University of Toronto. All rights reserved.
  Copyright 1996 Technion. All rights reserved.
  Copyright 1996 and 1997 University of Vermont. All rights reserved.
  Copyright 1997-2001 NEC Research Institute, Inc. All rights reserved.
  Copyright 2002-2013 Purdue University. All rights reserved.
  This program is free software: you can redistribute it and/or modify
  it under the terms of the GNU Lesser General Public License as published by
  the Free Software Foundation, either version 3 of the License, or
  (at your option) any later version.
  This program is distributed in the hope that it will be useful,
  but WITHOUT ANY WARRANTY; without even the implied warranty of
  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  GNU Lesser General Public License for more details.
  You should have received a copy of the GNU Lesser General Public License
  along with this program.  If not, see http://www.gnu.org/licenses.

chickadee » image-processing

Identifier search

Outdated egg!

TOC »