Title: Triangular fuzzy interpolation deinterlacing algorithm method for image edge detection
Authors: G.K. Jagatheswari; R. Murugesan
Addresses: Department of Science and Humanities, PES University, Electronic City Campus, Bangalore, 560 100, India ' Department of Physical Science, REVA University, Bangalore, 560 064, India
Abstract: Edge detection is the first step in image as it can be applied in the process of feature extraction, segmentation and classification. This study proposes a new fuzzy logic based edge detection method and computationally efficient algorithm based on triangular fuzzy interpolation technique for intra field deinterlacing of images that reduces the loss in image quality at the edges. The proposed algorithm improves image deinterlacing quality with respect to missing pixels in the edge region. When interpolating the missing pixels, characteristics or features of the region around the pixel are interpolated first. The proposed algorithm is tested with five test images which have both uniform surface and edges. The subjective and objective performance of the proposed algorithm shows greater improvement in the edge quality when correlated with the traditional deinterlacing methods. The designed fuzzy rules become an attractive solution for the betterment of the quality of edges as much as possible. The result reveals that the proposed algorithm provides greater robustness to contrast and lighting variations.
Keywords: triangular fuzzy representation; slope scale; missing pixel; edge detection; edge direction gradient; deinterlacing; fuzzy interpolation; Gaussian gradient; rough interlacing line.
DOI: 10.1504/IJFCM.2022.124338
International Journal of Fuzzy Computation and Modelling, 2022 Vol.4 No.1, pp.1 - 15
Received: 26 Jul 2019
Accepted: 25 Aug 2020
Published online: 25 Jul 2022 *