Font Size: a A A

Deep-sea Self-adaptive LED Supplementary Lighting System Based On Online Imaging Quality Analysis

Posted on:2024-04-20Degree:MasterType:Thesis
Country:ChinaCandidate:Y K JiangFull Text:PDF
GTID:2530307103972129Subject:Electronic information
Abstract/Summary:PDF Full Text Request
The deep-sea harbors vast mineral resources and diverse ecosystems,and the development and utilization of ocean resources have become a new field of technological competition among various countries.During deep sea exploration,due to the absorption,scattering,and attenuation of light by water,the lighting conditions are poor,and effective artificial lighting devices are essential equipment in deep sea exploration.To obtain good imaging effects,operators usually need to manually adjust the brightness of the fill light,which not only reduces the efficiency of exploration but also cannot be applied to autonomous exploration equipment(such as underwater autonomous submersibles).To improve the efficiency of deep-sea visual detection and the imaging effects of autonomous visual exploration equipment,this paper designs a deep-sea adaptive LED fill light system based on online image quality analysis and researches deep-sea lighting devices and adaptive fill light algorithms.To address the problem of the lack of effective lighting equipment during deep-sea exploration,this paper designs a deep-sea high-power LED lighting lamp based on the LT3797 driver chip.The circuit design,based on the BUCK-BOOST topology,allows the lamp to have a wide voltage input range,from 10 to 48 V.The maximum output luminous flux exceeds 8000 lm,with an output power of around 120 W.The lamp supports both PWM and RS485 dimming,and has functions such as overload protection and temperature monitoring.The design,which features titanium alloy housing and a sapphire glass window,allows the lamp to withstand a maximum pressure of 65 MPa.To address the problem of insufficient adaptive dimming methods in deep-sea environments,this paper proposes an adaptive LED lighting system based on online image quality analysis.First,a no-reference underwater image quality assessment method based on human visual characteristics is proposed to accurately evaluate the quality of underwater images.This method analyzes underwater image quality in real-time,extracting three indicators: chroma,sharpness,and contrast,to provide a comprehensive quality score.Experimental results demonstrate that this quality evaluation index has good accuracy,robustness,and consistency with subjective evaluation.Based on effective underwater image quality evaluation indicators,an adaptive lighting algorithm combining the simulated annealing algorithm and the greedy algorithm is proposed.The system automatically adjusts the lighting intensity of each light based on the optimal image quality and corresponding illumination intensity obtained by simulated annealing algorithm initialization.When the environment changes,the adaptive algorithm based on the greedy algorithm adjusts the light intensity of each light to obtain the local optimal solution of the current imaging quality,thereby improving the imaging quality and achieving better imaging results.The deep-sea lighting system designed in this paper has completed the verification tests of relevant technical specifications and lighting effects,and has undergone high and low-temperature experiments and high-pressure tests in a simulated deep-sea environment.The test results show that the lighting system can operate stably at temperatures ranging from-10℃ to 40℃ and can work normally under high pressure of 65 MPa,and all indicators meet the design requirements.The adaptive lighting system has also undergone relevant tests in an outdoor experimental water tank,and the experimental results show that the proposed adaptive lighting algorithm can adjust the brightness of the supplementary light according to the imaging screen,significantly improving the underwater imaging quality.
Keywords/Search Tags:Image quality evaluation, deep-sea lighting system, adaptive dimming, simulated annealing algorithm, greedy algorithm
PDF Full Text Request
Related items