Unsupervised Deep Contrast Enhancement with Power Constraint for OLED Displays

Area of Science:

• Display Technology
• Artificial Intelligence
• Image Processing

Background:

• Organic Light Emitting Diode (OLED) displays face challenges in balancing power consumption and image quality.
• Existing power-constrained contrast enhancement (PCCE) techniques often compromise visual fidelity.

Purpose of the Study:

• To propose a novel deep learning-based PCCE scheme for OLED displays.
• To enhance image contrast while constraining power consumption.
• To achieve this without requiring reference images through unsupervised learning.

Main Methods:

• A convolutional neural network (CNN) is employed to enhance image contrast.
• Power consumption is managed by reducing display brightness by a specific ratio.
• The CNN utilizes unsupervised learning to adapt the PCCE technique.

Main Results:

• The proposed deep learning method effectively enhances image contrast.
• Power consumption is successfully constrained while preserving perceived visual quality.
• Experimental results demonstrate superiority over conventional methods using metrics like VSI and EME.

Conclusions:

• The developed unsupervised CNN-based PCCE method offers a superior approach for OLED displays.
• This technique effectively balances power reduction and image quality enhancement.
• It presents a significant advancement in display technology optimization.