Switch machine notch monitoring system and method based on binocular visual processing identification

Abstract

The invention provides a switch machine notch monitoring system and method based on binocular visual processing identification, relating to the switch machine notch monitoring field. The detection system comprises an on-site binocular collection subsystem, a CAN bus fiber transmission subsystem and a monitoring subsystem. The binocular collection subsystem collects notch images, and transmits the images to the monitoring subsystem through the CAN bus fiber transmission subsystem; the monitoring subsystem performs visual processing, and employs a gauss filter to perform smooth filtering de-noising processing; a Canny edge detector extracts notch edge information, performs Hough transformation on edge images to detect line segments, and identifies line segment information in the images. Through a series of constraint conditions, effective line segments on the top left side of notch edges are screened, reference mark and real marking line segment coordinates are recorded, and notch offsets on the images are calculated according to image geometrical distances. According to proportions, real notch offsets are calculated. A left image and a right image are both processed according to the processing flow, thereby improving calculating accuracy.

Description

technical field [0001] The invention relates to a switch machine gap monitoring system and method, belonging to the field of turnout control in rail transit, in particular to a switch machine gap monitoring system and method based on binocular vision processing recognition. Background technique [0002] The safety and stability monitoring of the switch machine, the actuator of the turnout control system, plays an important role in the safe operation of trains in the rail transit industry. Gap status monitoring data (images and videos) is an important reference for analyzing the operating status of switch machines. [0003] Traditional switch machine gap monitoring methods are divided into two categories: direct measurement and indirect measurement. The sensor of the direct measurement method is installed in the gap, and its output can directly reflect the size of the measured gap, but the installation is complicated. Indirect measurement is easy to install, but requires ca...

AI Technical Summary

Problems solved by technology

The sensor of the direct measurement method is installed in the gap, and its output can directly reflect the size of the measured gap, but the installation is complicated

The indirect measurement method is easy to install, but requires careful adjustment of the reference position

The existing point machine monitoring system has the following problems: First, the electromagnetic measurement used in the existing system is greatly affected by temperature and electromagnetic, and some systems can only qualitatively reflect the state of the gap, and cannot perform quantitative analysis on the gap; Some systems also require different sensors for different switch machines

Second, the existing point machine gap monitoring systems and monitoring principles are diverse, and the various methods are incompatible, which increases the difficulty and cost of system operation and upgrading

Third, most of the existing monitoring systems use ADSL transmission, which brings a large delay to the real-time high-definition transmission of images and videos

Images