Monocular vision and binocular vision switching device for small robot

Abstract

The invention provides a monocular vision and binocular vision switching device for a small robot, comprising a wireless communication camera, a transmission assembly, a power assembly and a control assembly, wherein the transmission assembly is composed of a rotary motion and linear motion switching assembly and a guide rail, a camera fixing plate is fixed on a linear motion component of the transmission assembly, baffles are arranged at two sides of the transmission assembly, each baffle is provided with a trigger switch for the control assembly, the two ends of a rotary motion component of the transmission assembly are connected with the baffles by virtue of bearings, and one end of the rotary motion component penetrates through the baffles to be connected with the power assembly by virtue of a coupler; and the control assembly comprises a control circuit board as well as a first channel trigger switch and a second channel trigger switch which are respectively arranged on the baffles at the two sides of the transmission assembly. The monocular vision and binocular vision switching device provided by the invention realizes mutual switching between monocular vision and binocular vision, so that the small robot can be more widely applied; and binocular base line distance is adjustable, so that better rebuilding effect can be obtained by regulating base line distance.

Description

technical field [0001] The invention relates to the technical field of robot vision, in particular to a monocular vision and binocular vision conversion device for a small robot. Background technique [0002] The purpose of machine vision research is to enable the machine to have the ability to recognize three-dimensional environment information through two-dimensional image information. This ability not only enables the machine to perceive the geometric information (such as shape, position, posture movement, etc.) They can be further described, stored, identified and understood. At present, machine vision is mainly used for 3D reconstruction and distance calculation. In vision research, it is mostly divided into monocular stereo vision, binocular and multi-eye stereo vision, and currently binocular or monocular cameras are mostly used. [0003] The monocular vision method is a method of 3D reconstruction using one camera. The following two methods are often used: 1. Deri...

AI Technical Summary

Problems solved by technology

2. Match the same feature points in different images through multiple images from multiple viewpoints, and use these matching constraints to obtain spatial 3D point coordinate information, thereby realizing 3D reconstruction, which can realize camera self-calibration in the reconstruction process, and can meet large-scale scenes The demand for 3D reconstruction, and the reconstruction effect is better in the case of rich image resources; the disadvantage is that the calculation amount is relatively large, and the reconstruction time is long

But it has the following disadvantages: 1. The image formed by the object and its image overlaps, which brings trouble to the later image processing; 2. When the placement angle of the two sets of mirrors is changed, it is easy to cause the increase of the baseline distance and cause Increase the size of the stereo vision system; 3. The baseline distance is small and the measurement accuracy is low

This mechanism uses two cameras, which is costly and troublesome to install and maintain. At the same time, due to its large weight and volume, it cannot be used on small robots.

Images