Quantitative quality measure based on light wave distribution to access 3D display

Abstract

There are a lot of three-dimensional (3D) displaying methods such as stereoscopy, integral photography, holography, etc. These technologies have different 3D vision properties and 3D image qualities. Conventionally, biological responsiveness is measured by using an actual 3D display in order to evaluate image qualities of 3D displaying method. It is required quantitative quality measure for 3D images for quantitative evaluation, which are useful for comparing 3D image quality and a design of a new display system. In this paper, we propose quality measures for 3D images named volume signal to noise ratio (VSNR), which is a three-dimensionally extended signal to noise ratio (SNR). A 3D display produces light wave distributions in 3D space, which makes observers view 3D image illusions. The VSNR measures error of light wave distributions between generated by actual objects and produced by a 3D display. The light wave distribution is including various factors for 3D perception of human such as resolution of reconstructed images, visual fields, motion parallax, and depth of field. The VSNR evaluates these 3D perception factors totally. We were carried out the experiments to certificate the efficiency of the VSNR. 3D images represented electro-holographic display and integral photographic displays were evaluated by the VSNR. The results indicated that the electro-holographic display has better quality than integral photographic display, but speckle noise deteriorates the 3D image quality.