Simple Patch Matcher: Uma Abordagem Baseada em Amostras para Restauração de Imagens

• Gustavo Becelli do Nacimento ^[1] ; Almir Olivette Artero ^[1] ; Danilo Medeiros Eler ^[1] ; Maurício Araújo Dias ^[1] ; Danillo Roberto Pereira ^[1] ; Francisco Assis da Silva ^[2]
  1. [1] Universidade Estadual Paulista

     Universidade Estadual Paulista

     Brasil

     
  2. [2] Universidade do Oeste Paulista

     Universidade do Oeste Paulista

     Brasil

     
• Localización: RISTI: Revista Ibérica de Sistemas e Tecnologias de Informação, ISSN-e 1646-9895, Nº. 61, 2026, págs. 3-19
• Idioma: portugués
• Títulos paralelos:
  • Simple Patch Matcher: A Sample-Based Approach for Image Restoration
• Enlaces
  • Texto completo
• Resumen
  • español

    Despite the importance of digital photographs in preserving memories and cultural artifacts, the restoration of analog images remains a crucial task.

    However, this process continues to be a major challenge. Thus, this work proposes a technique called Simple Patch Matcher (SPM), which is based on the use of the Fast Marching Method (FMM) structure to select pixels, and sample similarity to restore pixels in degraded regions, using a multidimensional data structure to store non-degraded samples of the original image, which serve as candidates for pixel restoration in degraded areas. The results obtained with SPM demonstrate its effectiveness in restoration, surpassing conventional algorithms. This is evidenced by an increase of approximately 10% in the Peak Signal-to-Noise Ratio (PSNR) and significant improvements in other quality metrics, such as the Structural Similarity Index (SSIM) and the Root Mean Square Error (RMSE), compared to classical restoration methods.

  • português

    Apesar da importância das fotografias digitais na preservação de memórias e artefatos culturais, a restauração de imagens analógicas ainda é uma tarefa crucial. Porém, este processo continua sendo um grande desafio. Assim, este trabalho propõe uma técnica chamada Simple Patch Matcher (SPM), que se basea no uso da estrutura do Fast Marching Method (FMM) para selecionar os pixels, e similaridade de amostras para restaurar pixels em regiões degradadas, utilizando uma estrutura de dados multidimensional para armazenar amostras não-degradadas da imagem original, que servem como candidatos para a restauração dos pixels em áreas degradas. Os resultados obtidos com a técnica SPM demonstram sua eficácia na restauração, superando algoritmos convencionais. Isso é evidenciado pelo aumento de cerca de 10% na Relação Sinal Ruído de Pico (PSNR) e melhorias significativas em outras métricas de qualidade, como o Índice de Similaridade Estrutural (SSIM) e o Erro Médio Quadrático (EMQ), em comparação com técnicas clássicas de restauração.

• Referencias bibliográficas
  • Barnes, C., Shechtman, E., Finkelstein, A., Goldman, D. B. (2009). PatchMatch: A randomized correspondence algorithm for structural image...
  • Bertalmio, M., Bertozzi, A., Sapiro, G. (2001). Navier-stokes, fluid dynamics, and image and video inpainting, in Proceedings of the 2001...
  • Cheng, J., Li, Z. (2018). Markov random field-based image inpainting with direction structure distribution analysis for maintaining structure...
  • Conway, M. E. (1963). A multiprocessor system design. in Proceedings of the November 12-14, 1963, Fall Joint Computer Conference, AFIPS ’63,...
  • Elharrouss, O., Almaadeed, N., Al-Maadeed, S., Akbari, Y. (2019). Image inpainting: A review. Neural Processing Letters, 51:2007–2028. https://doi.org/10.48550/ arXiv.1909.06399.
  • Goodfellow, I., Bengio, Y., Courville, A. (2016). Deep Learning. MIT Press, http://www.deeplearningbook.org.
  • Ignatov, A., Timofte, R. (2019). Pirm challenge on perceptual image enhancement on smartphones: report. in European Conference on Computer...
  • Kuszmaul, C. L. (2023). Dictionary of Algorithms and Data Structures. https://www.nist.gov/dads/HTML/heap.html.
  • Liao, H., Funka-Lea, G., Zheng, Y., Luo, J., Zhou, S. K. (2018). Face completion with semantic knowledge and collaborative adversarial learning,...
  • Matsakis, N. Stone, J. (2023). Rayon: Simple work-stealing parallelism for Rust.
  • Oliveira, M., Bowen, B., McKenna, R., Chang, Y.-S. (2001). Fast digital image inpainting. In Proceedings of the International Conference on...
  • Bradski, G. (2000). The OpenCV Library. Dr. Dobb’s Journal of Software Tools.
  • Padua, D. (2011). Encyclopedia of Parallel Computing. Springer.
  • Rangel, J. J., Ruiz, J. S, García, F., & Cervantes, J. (2019). Redes Neuronales Convolucionales en la identificación de melanomas benignos...
  • Ritchin, F. (2009). After Photography. WW Norton & Company.
  • Rombach, R., Blattmann A., Lorenz, D., Esser, P., Ommer, B. (2021). Highresolution image synthesis with latent diffusion models, https://doi.org/10.48550/arXiv.2112.10752.
  • Shepard, D. (1968). A two-dimensional interpolation function for irregularlyspaced data.in Proceedings of the 1968 23rd ACM National Conference,...
  • Telea, A. (2004). An image inpainting technique based on the fast marching method.Journal of Graphics Tools, 9. https://doi.org/10.1080/ 10867651.2004.10487596.
  • You, S., You, N., Pan, M. (2019). Pi-rec: Progressive image reconstruction network withedge and color domain. arXiv preprint arXiv:1903.10146....
  • Wang, Z., Bovik, A., Sheikh, H., Simoncelli, E. (2004). Image quality assessment: from error visibility to structural similarity. IEEE Transactions...