FIELD: image recognition.
SUBSTANCE: at steps of method: for each image generating features vector consisting of brightness manifestation frequency distribution histogram values by levels gradations, in accordance with bit capacity; making a decision based on results of calculation of difference of feature vectors of recognizable image and each of reference images; wherein first from each distribution histogram, by its successive correlation with Haar wavelets having different scale parameters, three-dimensional matrix of amplitude coefficients with shift and scale coordinates is formed, and amplitude coefficients of the three-dimensional matrix averaged on the scale coordinate are selected as elements of the feature vectors.
EFFECT: high reliability of recognizing objects on images.
1 cl, 12 dwg
Title | Year | Author | Number |
---|---|---|---|
SIGNAL RECOGNITION METHOD | 2023 |
|
RU2809021C1 |
SIGNAL RECOGNITION METHOD | 2024 |
|
RU2831931C1 |
METHOD OF RADIO SIGNAL DISCRIMINATION | 2013 |
|
RU2533651C2 |
METHOD TO DETECT RADIO SIGNALS | 2011 |
|
RU2466455C1 |
METHOD OF IDENTIFYING RADIO SIGNALS | 2014 |
|
RU2551903C1 |
METHOD OF IDENTIFYING RADIO SIGNALS | 2007 |
|
RU2356064C2 |
METHOD OF DETECTING RADIO SIGNALS | 2010 |
|
RU2423735C1 |
RADIO SIGNAL RECOGNITION METHOD | 2010 |
|
RU2430417C1 |
RADIO SIGNAL RECOGNITION METHOD | 2010 |
|
RU2430416C1 |
METHOD FOR RECOGNITION OF RADIO SIGNALS | 2004 |
|
RU2261476C1 |
Authors
Dates
2025-05-06—Published
2024-07-02—Filed