Deep Drainage Detection System for Inland Vessels Based on Machine Vision

摘要

In order to prevent inland ships from overdraft, it is necessary to detect the draft depth of inland ships to ensure the safety of navigation. To design a river ship draught depth detection system based on machine vision, it collects accurate and comprehensive visual images of ships through image acquisition module, and provides industrial computer with reasonable detection methods to detect the draft depth of ships. The detection results are stored in the database and displayed in real time. The detection module is composed of FPGA and DSP hardware to realize the system detection process, communication and transaction, and the effective control of the terminal. When the inland water body is clear, the system uses the edge detection based draft depth detection method to effectively detect the ship draft depth. When the inland water body is turbid, the system uses binocular stereo vision three-dimensional detection method to measure the ship draft depth. The test results show that when the inland water body is clear, the error of the intake depth of inland ships detected by the detection system is ±0.05 m, and when the inland water body is turbid, the error of the system is within ±0.07 m. This shows that the detection system can accurately detect the intake depth of inland water body in clear and turbid conditions, and the test results are comprehensive and accurate.