Psychophysical and metric assessment of fused images

  • Authors:

  • Timothy D. Dixon;Jan Noyes;Tom Troscianko;Eduardo Fernández Canga;Dave Bull;Nishan Canagarajah

  • Affiliations:

  • University of Bristol, UK;University of Bristol, UK;University of Bristol, UK;University of Bristol, UK;University of Bristol, UK;University of Bristol, UK

  • Venue:
  • APGV '05 Proceedings of the 2nd symposium on Applied perception in graphics and visualization
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

The prevalence of image fusion - the fusing of images of different modalities, such as visible and infrared radiation - has increased the demand for accurate methods of image quality assessment. Two traditional methods of assessment that have been used are computational metrics and subjective quality assessment; we propose an alternative task-based method of image assessment, which represents a more accurate description of image 'quality' than subjective ratings. The current study used a signal detection paradigm, identifying the presence or absence of a target in briefly presented images followed by an energy mask, which was compared with computational metric results. In Experiment 1, 18 participants were presented with composites of fused infrared and visible light images, with a soldier either present or not. There were two independent variables, each with three levels: image fusion method (averaging, contrast pyramid, dual-tree complex wavelet transform), and JPEG2000 compression (no compression, low, and high compression), in a repeated measures design. Participants were presented with images and asked to state whether or not they detected the target. In addition, metric results were calculated and compared with task performance. Images were blocked by fusion type, with compression type randomised within blocks. This process was repeated in Experiment 2, but with JPEG images substituted for JPEG2000. The results showed a significant effect for fusion but not compression in JPEG2000 images, whilst JPEG images showed significant effects for both fusion and compression. The metric results for both JPEG and JPEG2000 showed similar trends with more advanced metrics matching the performance of the psychophysical tests more accurately.