Documentaton - BLOCKS - BLOCK_HIST

This block extracts histograms from images in a database.

BK = BLOCK_HIST() Initializes the block with the default options.

BK = BLOCK_HIST(BK) Executes the block with options and inputs BK.

Required Inputs:

db

The database of images.

feat

Features extracted from the database.

dict

A dictionary to use on the features.

Options:

bk.min_sigma

Discard features with scale less than min_sigma. Default 0.

bk.max_num

Keep at most max_num features, selected randomly. Default +inf.

bk.ref_size

ref_size used in feature extraction. Required to accurately measure min_sigma. Default [] does not modify the scale of the features.

bk.rand_seed

Set the random seeds before proceeding. Default [] does not modify the random seeds.

Fetchable attributes:

db

The database used.

dictionary

The dictionary used.

histogram

The histogram of a particular image. Required argument: seg_id.

words

The quantized words which form the histogram. Required argument: seg_id.