hlac: Higher Order Local Autocorrelation (HLAC)

In wvtool: Image Tools for Automated Wood Identification

Description

Feature extraction for practical vision system, whose features are shift-invariant and additive. The function gives zero to the eighth order cases, represented by 223 mask patterns of 3 x 3 within a 2r+1 x 2r+1 (r >=1) displacement region.

Usage

hlac(x, r=1, disp=FALSE)

Arguments

x	A binary or gray image or matrix
r	Displacement vector r for 3 x 3 mask pattern
disp	If TRUE, function saves 223 filtered images in one matrix.

Details

The feature parameter should be a list. The function returns 1, 4, 20, 45, 62, 54, 28, 8,1 features and corresponding filtered images if disp is TRUE.

Value

HLAC features or the corresponding image with requested HLAC measures.

features	Numerical output of 0 to 8th order masks
mat	A large matrix of 223 images expanded in a row

References

N.Otsu, T. Kurita (1988) A New Scheme for Practical Flexible and Intelligent Vision Systems, In: Proc. Machine Vision Application(MVA), 431-435.

See Also

rgb2gray, gray2bin, glcm, lbp

Examples

# features plot and the corresponding image presentation
data(camphora)
tmp <- hlac(gray2bin(camphora), 2, disp=TRUE)
par(mfrow=c(2,2))
plot(unlist(tmp$features), main="HLAC histogram")
image(rot90c(matrix(tmp$mat[2,],tmp$row,tmp$col)),
	col  = gray((255:0)/255), main="2", useRaster=TRUE, asp=1, axes=FALSE)
image(rot90c(matrix(tmp$mat[23,],tmp$row,tmp$col)),
	col  = gray((255:0)/255), main="23", useRaster=TRUE, asp=1, axes=FALSE)
image(rot90c(matrix(tmp$mat[156,],tmp$row,tmp$col)),
	col  = gray((255:0)/255), main="156", useRaster=TRUE, asp=1, axes=FALSE)

wvtool documentation built on May 1, 2019, 10:27 p.m.