Abstract
Motion estimation is a low-level vision task that is especially relevant due to its wide range of applications in the real world. Many of the best motion estimation algorithms include some of the features that are found in mammalians, which would demand huge computational resources and therefore are not usually available in real-time. In this paper we present a novel bioinspired sensor based on the synergy between optical flow and orthogonal variant moments. The bioinspired sensor has been designed for Very Large Scale Integration (VLSI) using properties of the mammalian cortical motion pathway. This sensor combines low-level primitives (optical flow and image moments) in order to produce a mid-level vision abstraction layer. The results are described trough experiments showing the validity of the proposed system and an analysis of the computational resources and performance of the applied algorithms.
Highlights
There are several definitions of the goal of visual perception [1,2] as the interpretation of the information arriving at the retina, while a general agreement about the different abstraction levels and the limits between them is lacking.Low-level vision obtains useful measurements such as colour, spatial frequency, binocular disparity, motion processing, etc., from several channels
Mid-level vision integrates primitives processed at a previous level. Information delivered at this stage corresponds to real-world inferences such as egomotion and independent moving objects (IMOs)
We present a prototype based on a FPGA device suitable for industrial applications which involves reduced size, rapid prototyping, low cost and power consumption
Summary
There are several definitions of the goal of visual perception [1,2] as the interpretation of the information arriving at the retina, while a general agreement about the different abstraction levels and the limits between them is lacking. Low-level vision obtains useful measurements such as colour, spatial frequency, binocular disparity, motion processing, etc., from several channels Some of these channels, or space-temporal filters, can be identified with receptive fields that deliver information to the retina. Mid-level vision integrates primitives processed at a previous level Information delivered at this stage corresponds to real-world inferences such as egomotion and independent moving objects (IMOs). Our bioinspired sensor integrates two Low-level vision primitives represented by gradient family optical flow estimation and variant orthogonal moments. Orthogonal variant moments improve the robustness of the final system featuring the pixels Both early-vision cues provide information for the Mid-Level output which has been configured in this contribution in the framework of segmentation and tracking tasks.
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call