Research on Gamma Correction of Field-stacked Silicon-based OLED Micro-displays Based on Genetic Algorithm
In the context of the emerging and evolving concept of the Metaverse,the technology of Virtual Reality(VR)has incrementally unveiled its substantial importance.Within this particular domain,micro-displays,functioning as a pivotal interface bridging the virtual world with the tangible reality,have assumed an immensely crucial role.Silicon-based OLED micro-displays,in particular,distinguished by their attributes of high resolution,superior contrast,and exceptionally vivid colors,have emerged as cornerstone products in the arena of next-generation micro-display technologies.The operational methodologies for these silicon-based OLED micro-displays are primarily bifurcated into two types:digital driving and analog driving.Digital driving,acclaimed for its prompt response and elevated contrast levels,has been extensively embraced in the industry.This specific mode of operation predominantly utilizes the technique known as Pulse Width Modulation(PWM),a method employed to generate an array of distinct grayscale levels.This is achieved by meticulously adjusting the proportional duration of pixel activation and deactivation.Within the diverse landscape of PWM methodologies,the technique of field-stacked driving is particularly noteworthy.This method ingeniously orchestrates fields with varying weights in a meticulously structured sequence,effectively diminishing the instantaneous bandwidth while proficiently representing diverse levels of grayscale.Nevertheless,one can not overlook the significance of the brightness pulses that are generated by the equivalent capacitive characteristics of OLED devices during their activation phase.In the scenario of field-stacked driving,the brightness pulses emanating from fixed fractional subfields that undergo on-off transitions have a direct and profound impact on the displayed brightness,consequently leading to a nonlinear escalation in the grayscale curve.This issue predominantly manifests in two forms:the nonlinearity of the grayscale curve itself,and a paradoxical decrease in brightness as the grayscale increases,culminating in the emergence of ineffective grayscale points.Together,these challenges add a layer of complexity to the process of Gamma correction.A prevalent strategy in Gamma correction is the augmentation of bit depth,which offers a broader spectrum of grayscale levels,thereby allowing for a more precise approximation of the nonlinear characteristics inherent to the Gamma 2.2 curve.A linear progression in the grayscale curve simplifies the Gamma correction process by obviating the need for individual point adjustments.However,the characteristic of nonlinear progression in the grayscale curve leads to a reduction in the quantity of utilizable grayscale levels,thereby impinging upon its linear portrayal.To execute Gamma correction effectively,it is imperative to eradicate the nonlinear progression present within the grayscale curve.In response to this necessity,this paper introduces an innovative brightness model.This model is founded on the principles of non-ideal field-stacked digital driving and incorporates a synthesis of various elements such as the sequencing of fields,the weighting of fields,the Vcom voltage value,and the configuration of Vcom voltage.This integration effectively reconstructs the grayscale curve that has been impacted by non-ideal brightness pulses.By judiciously adjusting these parameters,it becomes feasible to substantially diminish the frequency of brightness pulse occurrences and to compensate for the impacts of non-ideal brightness pulses.Consequently,this paper employs a genetic algorithm to optimize the grayscale curve,with the explicit objective of minimizing the root mean square error and the count of ineffective grayscale points between the actual grayscale curve and its ideal counterpart.This model can be calibrated using a nonlinear least squares fitting approach by measuring various Vcom values and time t,along with corresponding brightness levels,on a full-color silicon-based OLED micro-display with a resolution of 2 560×2 560×3.By applying the non-ideal field-stacked driven OLED brightness model in conjunction with a genetic algorithm,this paper meticulously develops appropriate populations and fitness functions specifically designed to optimize both the root mean square error and the number of ineffective grayscale points in the grayscale curve.Through the iterative process of evolving multiple generations of these populations,the paper successfully identifies the optimal population,which effectively represents the most advantageous parameter space.When this optimal parameter space is implemented and measured in a full-color silicon-based OLED micro-display,a marked improvement in the grayscale curve is observed.The optimization process significantly reduces the root mean square error from an unoptimized state of 21.65 cd/m2 and 15 395 redundant grayscale points to a much more refined state of 1.62 cd/m2 and a mere 2 977 points.The Gamma 2.2 curve,post-optimization,successfully aligns with the ideal characteristics of the Gamma 2.2 curve,and it exhibits a notably enhanced differentiation in the low grayscale range,especially when compared to traditional analog driving techniques.
NETL
NSTL
万方数据
