Stereoscopic Optical Correlator for Depth Estimation

Abstract

The stereo vision technique is capable of extracting depth information. Therefore, it is used in various areas, including robotics, augmented reality, autonomous vehicles, medical imaging, and space applications. Although the technique has evolved into a potent tool, time consumption remains one of the primary concerns in its utility. On the other hand, pattern recognition with optical correlators is invaluable when fast processing is required. To devise a computationally effective method, in this letter, we combine the information processing capability of optical technology with stereo vision. We develop a stereoscopic optical correlator that exploits the optical correlator’s fast processing capability and stereo vision technique’s proficiency to address the concern of processing time and power consumption. To test the ability of the proposed stereoscopic optical correlator, the processor has been applied to the prepared datasets, and good accuracy in depth estimation has been achieved.