A local dimming method based on improved multi-objective evolutionary algorithm

The advent of the digital era dramatically enriches the information obtained by people through vision and urges people to put forward higher requirements for image display quality. The luminance dynamic range is an essential factor affecting the image display quality. In recent years, the local dimming system whose backlight module consists of Light Emitting Diode (LED) backlight blocks is proposed. By setting the luminance of each backlight block separately, the local dimming system can effectively enlarge the luminance dynamic range and improve the display quality. However, most of the existing local dimming methods only aim to improve the display quality or reduce the system power consumption. To balance the relationship between image quality and power consumption of the display system, this paper builds a multi-objective optimization model for local dimming tasks with the optimization objectives of reducing image distortion, improving image contrast ratio and reducing system power consumption. In order to effectively solve the optimization model, this paper enhances the Decomposition-based Multi-Objective Evolutionary Algorithm (MOEA/D) and proposes an improved multi objective evolutionary algorithm combining MOEA/D with Shuffle Frog Leaping Algorithm (SFLA), named MOEA/D-SFLA. The experimental results show that MOEA/D-SFLA can reduce the image distortion, improve the image contrast and reduce the system power consumption compared with the traditional local dimming methods. It is also proved that the MOEA/D-SFLA has better searching ability compared with MOEA/D and two improved MOEA/Ds.