Skip to main content

Table 2 Reconstruction quality against ground truth for the examined line sequences

From: Blotch and scratch removal in archived film using a semi-transparent corruption model and a ground-truth generation technique

System

LabL1

DanceL1

DanceL2

DanceL3

Furnace

    

Average

0.92 ±0.04

0.85 ±0.05

0.88 ±0.03

0.92 ±0.03

Minimum

0.85

0.75

0.83

0.85

Maximum

0.97

0.92

0.93

0.95

JOMBEI

    

Average

0.95 ±0.01

0.92 ±0.02

0.91 ±0.02

0.95 ±0.02

Minimum

0.93

0.88

0.87

0.89

Maximum

0.97

0.96

0.96

0.96

BTBR-AOFF

    

Average

0.94 ±0.02

0.90 ±0.03

0.91 ±0.03

0.92 ±0.03

Minimum

0.87

0.85

0.88

0.85

Maximum

0.97

0.96

0.96

0.96

BBM

    

Average

0.94 ±0.03

0.91 ±0.02

0.90 ±0.02

0.92 ±0.02

Minimum

0.88

0.85

0.85

0.89

Maximum

0.98

0.84

0.9

0.94

BTLR

    

Average

0.96 ±0.02

0.92 ±0.02

0.94 ±0.02

0.95 ±0.02

Minimum

0.93

0.87

0.90

0.90

Maximum

0.98

0.96

0.96

0.96

  1. SSIM [50] is used here where an SSIM of 1 means that the sequence is identical to the ground truth reconstruction. The average and the minimum SSIM show that BTLR is the best performing technique.