Image-Quality-Based Power Control Technique for Organic Light Emitting Diode Displays

Abstract

This paper presents a new image-quality-based power control algorithm for an organic light-emitting diode display. The proposed algorithm calculates the optimal tradeoff between image quality and power consumption by employing an objective function that integrates a roll-off curve to eliminate the clipping error, which severely degrades the perceptual image quality. In addition, a sampling technique is adopted to reduce the computational complexity. The simulation results show that the proposed algorithm reduces the average power consumption by up to 0.310 (11.75% decrease) while preserving the image quality above the target level. When applying the sampling technique to the proposed method, the average computation time per pixel is reduced by up to 1.025 s (37.63% decrease) whereas the image quality and power consumption remain virtually unchanged.Index Terms-Clipping error control, organic light-emitting diode, power control.

Low-Overhead Adaptive Brightness Scaling for Energy Reduction in OLED Displays

“…The most suitable image transformation for real-time usage is brightness scaling, which is therefore the common solution adopted in consumer devices nowadays [16], [24], [25]. For grayscale images, brightness scaling transforms every pixel with the function:…”

“…However, as mentioned in Section 1, brightness scaling is generally image-agnostic, and uses the same scaling factor for all images. The authors of [25] and [24] first investigated adaptive brightness scaling approaches. In both works, the proposed transformations minimize power consumption under a minimum quality constraint, using a different scaling factor for each image.…”

“…In both works, the proposed transformations minimize power consumption under a minimum quality constraint, using a different scaling factor for each image. Specifically, in [25] only brightness values higher than a minimum knee-point are scaled, and the acceptable quality reduction is quantified in terms of maximum Mean Squared Error (MSE) between input and output images. The work in [24], conversely, uses a uniform scaling for all brightness values, and its quality constraint is specified as a minimum threshold on the Mean Structural Similarity Index (MSSIM) [17] between input and output, which correlates better with the subjective perception of image alteration compared to the MSE.…”

“…Although effective, these methods are still not suitable for an implementation in real time. Indeed, the whole transformation process takes several seconds for a 1080p image [24], [25].…”

Unsupervised Deep Contrast Enhancement With Power Constraint for OLED Displays

“…As an OLED directly emits light at each pixel, the total power consumption of the display is known to be proportional to the sum of the power consumption of each pixel [3]- [5]. However, as the brightness of the overall pixels is reduced to save power, the visibility and contrast of the image with re- 1 duced brightness are also decreased. To cope with this problem, Lee et al [3] proposed a power-constrained contrast enhancement (PCCE) technique that saves power while enhancing the contrast of the image to preserve visual quality.…”