Visual masking is an effect that contents of the image reduce the detectability of a given target signal hidden in
the image. The effect of visual masking has found its application in numerous image processing and vision tasks.
In the past few decades, numerous research has been conducted on visual masking based on models optimized
for artificial targets placed upon unnatural masks. Over the years, there is a tendency to apply masking model
to predict natural image quality and detection threshold of distortion presented in natural images. However, to
our knowledge few studies have been conducted to understand the generalizability of masking model to different
types of distortion presented in natural images. In this work, we measure the ability of natural image patches
in masking three different types of distortion, and analyse the performance of conventional gain control model
in predicting the distortion detection threshold. We then propose a new masking model, where detail loss and
additive defects are modeled in two parallel vision channels and interact with each other via a cross masking
mechanism. We show that the proposed cross masking model has better adaptability to various image structures
and distortions in natural scenes.
Proc. SPIE. 9016, Image Quality and System Performance XI
KEYWORDS: Target detection, Visual process modeling, Detection and tracking algorithms, Data modeling, Visualization, Image processing, Image quality, Image filtering, Performance modeling, RGB color model
Masking is a perceptual effect in which contents of the image reduce the ability of the observer to see the target signals hidden in the image. Characterization of masking effects plays an important role in modern image quality assessment (IQA) algorithms. In this work, we attribute the reduced sensitivity to the inhibition imposed by adjacent visual channels. In our model, each visual channel is excited by the contrast difference between the reference and distorted image in the corresponding channel and suppressed by the activities of the mask in adjacent channels. The model parameters are fitted to the results of a psychophysical experiment conducted with a set of different natural texture masks. Cross-validation is performed to demonstrate the model's performance in predicting the target detection threshold. The results of this work could be applied to improve the performance of current HVS-based IQA algorithms.