Template Matching with Noisy Patches: A Contrast-Invariant GLR Test

Charles-Alban Deledalle (1) , Loïc Denis (2) , Florence Tupin (3)

1 IMB - Institut de Mathématiques de Bordeaux
2 LHC - Laboratoire Hubert Curien
3 LTCI - Laboratoire Traitement et Communication de l'Information

Matching patches from a noisy image to atoms in a dictionary of patches is a key ingredient to many techniques in image processing and computer vision. By representing with a single atom all patches that are identical up to a radiometric transformation, dictionary size can be kept small, thereby retaining good computational efficiency. Identification of the atom in best match with a given noisy patch then requires a contrast-invariant criterion. In the light of detection theory, we propose a new criterion that ensures contrast invariance and robustness to noise. We discuss its theoretical grounding and assess its performance under Gaussian, gamma and Poisson noises.

Mots clés: Template matching; Likelihood ratio test; Detection theory; Image restoration

Domaines: Traitement des images [eess.IV]; Traitement du signal et de l'image [eess.SP]; Applications [stat.AP]; Traitement du signal et de l'image [eess.SP]

Charles-Alban Deledalle, Loïc Denis, Florence Tupin. Template Matching with Noisy Patches: A Contrast-Invariant GLR Test. European Signal Processing Conference 2013, Sep 2013, Marrakech, Morocco. ⟨hal-00804483⟩