Identifikasi Objek Menggunakan Proses Deteksi Tepi Metode Laplacian of Gaussian Dan Canny Terhadap Citra Sidik Jari
Autor: | Edi Suharto, Muhammad Yasin Simargolang, Muhammad Noor Hasan Siregar, Agus Perdana Windarto |
---|---|
Rok vydání: | 2022 |
Zdroj: | JURNAL MEDIA INFORMATIKA BUDIDARMA. 6:294 |
ISSN: | 2548-8368 2614-5278 |
DOI: | 10.30865/mib.v6i1.3459 |
Popis: | Identification is the identification or determination of an object based on evidence as a clue. The objective of the research was to identify biometric images using edge detection of LoG (Laplacian of Gaussian), Canny, and LoG+Canny with different shapes and dimensions. It is expected that the object can still be identified with different shapes and dimensions. The sample of data used was 20 fingerprint images. This fingerprint image was tested using the methods LoG, Canny and LoG+Canny. The process begins with the image reading, and then the image is converted to grayscale, edge detection and image segmentation. The final result is the identification of the image. The results show that the average accuracy is 89.9 per cent for the LoG method, while 81.8 per cent for the Canny method and 90.7 per cent for the LoG + Canny method. From 10 fingerprint image tests, 8 fingerprint images can be identified by both methods. While the LoG + Canny method is capable of identifying 9 fingerprint images. The LoG method can detect images of 2, 4, 5, 6, 7, 8, 9, 10; while the Canny method can detect images of 2, 3, 4, 6, 7, 8, 9, 10; and the LoG + Canny method can detect images of 1, 2, 3, 4, 6, 7, 8, 9, 10. The minimum and maximum pixel values for the LoG method are 11 pixels for the test image and 25327 pixels for the database image. While the minimum and maximum pixel values for the Canny method are 148 pixels for the test image and 42323 pixels for the database image. In the meantime, the minimum and maximum pixel values for the LoG + Canny method are 806 pixels for the test image and 57972 pixels for the database image. The LoG + Canny method can outperform other methods for the identification of fingerprint images from the results of the tests carried out. In addition to the higher accuracy value, the resulting error value is also much smaller. The object images in the LoG method that have not been identified are numbers 1 and 3 with an error of 27.27 percent and 58.33. While the Canny method that has not been identified is number 1 and 5 with an error of 98.31 per cent and 59.92 per cent. The LoG + Canny method that cannot be identified is number 5 with an error of 61.69 per cent. The mean error values for the three methods were 10.1%, 18.2% and 9.3% (LoG, Canny, LoG + Canny). |
Databáze: | OpenAIRE |
Externí odkaz: |