自适应直方图均衡化
峰值信噪比
均方误差
计算机科学
直方图
人工智能
图像质量
计算机视觉
过程(计算)
虚拟现实
图像(数学)
数学
直方图均衡化
统计
操作系统
作者
I Gede Partha Sindu,Rukmi Sari Hartati,Made Sudarma,Nyoman Gunantara
标识
DOI:10.1109/icitacee58587.2023.10277512
摘要
The research aims to enhance immersive VR through improved design display quality in virtual environments. It is grounded in the image problem because in the virtual environment, there is a difference in color contrast 3D design interior and exterior environments of residential with a real view of the interior and external environments of buildings. So the results of the design of the interior and exterior environments of housing are still not immersive and do not look natural or real. The method used is image enhancement consisting of Histogram Enhancement (HE), Contrast-Limited Adaptive Histogram (CLAHE), and Fuzzy Contrast Enhancement (FCE). The data used in this study totaled 3.157 images in PNG format with a resolution of 512 x 512. The stages of this study start with data finalization, the image enhancement process, and evaluation. Through the investigative process, CLAHE method is able to outperform HE and FCE with an average of Structural Similarity Index (SSIM) 0.959, Image Quality Index (IQI) 0.959, through Mean Squared Error (MSE) 0.003, Root Mean Squared Error (RMSE) 0.055, and Peak Signal-to-Noise Ratio (PSNR) 25.705. Additional parameters, such as tiles generation and clip limit, on CLAHE can improve the image quality of the virtual environment without causing over-enhancement.
科研通智能强力驱动
Strongly Powered by AbleSci AI