A FINGERPRINT REMOVAL METHOD BASED ON FRACTAL–CRIMINISI TECHNOLOGY
| Autor: | ZUN LI, AIMIN CHEN, TONGJUN MIAO |
|---|---|
| Rok vydání: | 2022 |
| Předmět: | |
| Zdroj: | Fractals. 30 |
| ISSN: | 1793-6543 0218-348X |
| DOI: | 10.1142/s0218348x22501572 |
| Popis: | Against the circumstance of redundant fingerprint information remaining in photographed images due to lens contamination, a fingerprint removal method is proposed by combining fractal technology and the Criminisi method. Fingerprints on an image are usually distributed in a dispersive pattern, and show statistical fractal characteristic. Therefore, based on the Criminisi method, the textural complexity of image pixel blocks expressed by box-counting fractal dimension is introduced into this method as a parameter of pixel block size and subitem of the data item with priority. On the one hand, this way of operation can select the size of the pixel blocks to be removed adaptive to the box fractal dimension; on the other hand, the box fractal dimension, as the subitem of the data item with priority, can avoid the effect of particular value and improve the reliability and stability of priority. The experimental results show that the proposed method can satisfactorily perform the fingerprint removal task and fill the information. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|