Triangular Similarities of Facial Features to Determine: The Relationships Among Family Members
Autor: | Y B Ravi Kumar, C. K. Narayanappa |
---|---|
Rok vydání: | 2018 |
Předmět: | |
Zdroj: | Journal of Advanced Computational Intelligence and Intelligent Informatics. 22:323-332 |
ISSN: | 1883-8014 1343-0130 |
DOI: | 10.20965/jaciii.2018.p0323 |
Popis: | An image can be represented in the form of patterns of intensities, with the objects of an image appearing in the form of a pattern on an X-Y plane. The two patterns of intensities of two corresponding facial images are measured by calculating the areas of right triangles formed from patterns in a Cartesian coordinate system. The purpose of representing patterns of intensities in the Cartesian coordinate system is to measure the percentage of similarities that exists between two facial images, similarities inherent in photographs. The percentage is measured by incorporating the proposed technique of areas that are common between two patterns of intensities. The pattern 1 produces areas of right triangles of a parent with respect to areas of right triangles of a child. The strategy of measuring the facial similarities between two patterns of intensities is dependent on the areas of pattern 1 that have commonalities with the areas of pattern 2. This helps in the measuring of the facial similarities between two patterns of intensities. The proposed method has yielded results of 71.3, 77.1, 71.3, and 70.5 percent of similarity on the dataset KinfaceW-I and 80.7, 82.1, 80.6, 81.1 on the dataset KinfaceW-II. |
Databáze: | OpenAIRE |
Externí odkaz: |