基于自适应光晕抑制算法的车载分区背光显示系统的研究及设计

打开文本图片集

键词：自适应光晕抑制算法；车载显示;Mini-LED；光晕效应；FPGA 中图分类号：TN873;TN911 文献标识码：A doi：1O.37188/CJLCD.2025-0115 CSTR：32172.14.CJLCD.2025-0115

Research and design of an onboard zoned backlight display system based on an adaptive halo suppression algorithm

LU Jiale，SUI Jixiang*，HUANG Haibo，HURui，CAO Yuping，HU Ao （School of Artificial Intelligence， Hubei University ofAutomotive Technology，Shiyan 442OOo，China）

Abstract：Mini-LED local dimming liquid crystal display （LCD） systems have emerged as a primary development direction for in-vehicle HDR displays due to their exceptional local dimming capability and ultra-high contrast ratio. However， in complex driving scenarios such as rapid tunnel traversal and dynamic nighttime lighting，existing backlight extraction algorithms like fixed-weight coefficient error correction suffer from halo difusion and nonlinear luminance distortion. This project conducted human visual perception experiments under varying ambient light conditions and iluminance levels. Two-factor ANOVA revealed that ambient light dominantly modulates halo perception. Based on Sigmoid and exponential decay functions，a halo index-ambient illminance-luminance diference model was established， deriving the relationship between the halo index ∇[H] and ambient illuminance （E） . After segmental analysis and simplification，the relationship between the correction value （ |C⟩ andambient illuminance（ ⋅E ） was obtained.Building on this，an adaptive halo suppression algorithm for partitioned backlight display systems was designed. A U-Net-based deep learning model was constructed for comparative simulation with the proposed algorithm. Finally，the algorithm was implemented and tested on an FPGA system-on-chip （SoC）.Experimental results indicate that while the halo control performance of the adaptive algorithm is marginally inferior to the deep learning approach，it significantly outperforms in processing speed. The adaptive algorithm achieves a unit-time gain （PSNR/inference time） of 280dB/s ，fivefold higher than U-Net （ 56dB/s ）. This“stable high brightness with rapid processing”characteristic ensures superior user visual experience in critical scenarios like in-vehicle navigation and rearview imaging.

Key Words： adaptive halo suppression algorithm； in-vehicle display； Mini-LED； halo effect； FPGA

1引言

随着汽车科技的飞速发展，车载显示系统在提升驾驶安全和乘车体验中扮演着日益重要的角色，而Mini-LED技术凭借高亮度、高对比度及成本优势正逐渐成为未来车载显示的发展方向[1-2]。（剩余12644字）