SPEC Benchmark Description File

Benchmark Name


Benchmark Author

ImageMagick Studio LLC, a non-profit organization dedicated to making software imaging solutions freely available.

Benchmark Program General Category

Image Manipulation.

Benchmark Description

ImageMagick is a software suite to create, edit, compose, or convert bitmap images. It can read and write images in a variety of formats (over 100) including DPX, EXR, GIF, JPEG, JPEG-2000, PDF, PhotoCD, PNG, Postscript, SVG, and TIFF. Use ImageMagick to resize, flip, mirror, rotate, distort, shear and transform images, adjust image colors, apply various special effects, or draw text, lines, polygons, ellipses and Bézier curves.

Input Description

The ImagMagick program in the SPEC benchmark configuration takes an input, manipulates it and produces a transformed image.

For the reference datasets, three different operations are done. For the first, a multicolor dot is the input. A shear=31 function is applied and the image is resized to 12000x9000 and then negated. An edge=14 operation is applied followed by implode=1.2 and then flop. This is followed by a convolve=1,2,1,4,3,4,1,2,1 operation, then and edge=100 operation and finally the image is resized to 800x600. For the second, a 900x900 checkerboard in input. Then a shear=31x14 function is applied, followed by negate then edge=14. The image is then resized to 12288x12288, followed by implode=1.2 function, flop function, and convole=1,2,1,4,3,4,1,2,1 function. The image then has edge=100 applied and resized down to 900x900 and written out. For the last, the same 900x900 checkerboard is input and resized to 12000x12000 and have a 5x5 convolve operator applied. It then as edge=60 applied, followed by log convolve operator, negate operation, followed by laplacian convolve function. A sharpen function=10x0.2-32 is applied followed by a resize to 600x600 for output.

For the train dataset, a small blue dot is the input. A shear=31 function is applied, the image is resized to 1280x960 and is then negated. The edge=14 function is applied, followed by implode=1.2. The image then has flop applied, followed by a convole=1,2,1,4,3,4,1,2,1 operation. Finally an edge=100 operation is applied before writing it out.

For the test dataset, a small blue dot is the input. A shear=25 function is applied, the image is then resizes to 640x480, it is then negated and then alpha=off operation is applied.

Output Description

Validation is accomplished by comparing the output image with an expected output. The comparison program then assigns a weight on how close the images are to being the same. This value is the basis for determined if the results look the same or not.

For the reference dataset, 3 different outputs are create, a 800x600 image, a 900x900 image and a 600x600 image. For the train dataset, a 1280x960 output image is created. For the test dataset, a 640x480 output image is created.

Programming Language


Known portability issues



Image Magick Homepage

Last Updated: 7 June 2012