TESLA
MMX Image Processing Library
High
speed image processing from Inventions
Tesla is Inventions’ new high speed MMX image processing library. It is faster than standard 32 bit RGB Windows MMX imaging software for colour images, as it processes more pixels in each instruction. Image formats include 8 and 16 bit mono, 24 bit colour and 32 bit Windows bitmaps. Tesla can be used with any frame-grabber that is compatible with Windows NT/95/98. It includes numerous fast functions for image arithmetic, thresholding, convolution, texture analysis, edge enhancement, smoothing, morphology and line detection. Tesla can be used in combination with Inventions’ ILIB imaging libraries or their Edison image processing package.
PRICE: £ 250 / US$ 425
For further details of Tesla or other Inventions imaging software,
please contact:
TESLA imaging functions
Image addition; add value to image; image subtraction; subtract value from image; threshold image; make binary image; logical AND; logical OR; multiply images; multiply image by value; rectangular mean filter; mean filter; weighted mean filter; horizontal edge filter; vertical edge filter; two crisp filters.
Prewitt row gradient, column gradient, North, North East, East, South East, South, South West, West, and North West filters; Robinson 3 North, North East, East, South East, South, South West, West, and North West filters; Robinson 5 North, North East, East, South East, South, South West, West, and North West filters; Kirsch North, North East, East, South East, South, South West, West, and North West filters.
Sobel row gradient, Sobel column gradient, Laplacian 4, Laplacian 8, Laplacian column, Laplacian row and separable Laplacian filters, pixel difference row and column filters, Roberts North East and North West filters, 9 Chebyshev texture filters; North South, East West, North South East West; North West South East line filters, North South, East West, North South East West, North West South East Weighted line filters, 9 Laws texture filters.
4 and 8 connected dilate; general 3x3 dilate; general NxM dilate; 4 and 8 connected erosion; general 3x3 erosion; general NxM erosion; image logical complement; outlier removal; spot detection filter; bit plane slice, multiband colour threshold; 3x3 integer convolve; 3x3 integer convolve with shift; 5x5 integer convolve; 5x5 integer convolve with shift; ; 7x7 integer convolve; 7x7 integer convolve with shift; 9x9 integer convolve; 9x9 integer convolve with shift NxM integer convolve; NxM integer convolve with shift.
Zero image; constant image; bit shift pixels right, bit shift pixels left, logical difference of two images, general hit and miss filter; 12 specific kernel hit and miss filters, 4 and 8 connected open; generalised 3x3 open; generalised NxM open; 4 and 8 connected close; generalised 3x3 close; generalised NxM close; 4 and 8 connected boundary extraction.
Region filling and region filling to limit; image comparison; connected image regions and connected image regions to limit; convex hull.
Thinning and thinning to limit; morphological 4 connected and diagonal pruning; thicken and thicken to limit; isolated pixel removal.