Research Of LED Dynamic Backlight Algorithm And Backlight Driven Circuit Control Method

Liquid Crystal Display (LCD) has been increasingly used in displays because it has many good features such as thin, light, high resolution, long life and good color performance. Therefore, it plays a dominant position both in large-size Television display domain and small-size mobile devices. However, since the LCD is not self-luminance display device, its working principle is using electrical stimulation of the liquid crystal molecules to generate point, line and surface image with the LCD back light source. Then, LCD requires backlight unit. Conventionally, LCD display system uses Cold Cathode Fluorescent Lamp (CCFL) as the backlight light source. However, there are two major drawbacks that are large power consumption and low display contrast ratio due to the light leakage. In recent year, lighting-emitting diode (LED) gradually replaces the CCFL as the backlight light source due to its low power consumption and can be dynamically controlled to enhance the image contrast. In this paper, our study focus on the LED dynamic backlight both in the mobile devices and large-size LCD TVs and we propose a novel multi-blocks light guide plate, Clipping Ratio Threshold backlight luminance control method and ranks decomposition backlight driven circuit control method.Firstly, the LED backlight structure of LCD display system has two major types that are direct type and edge type. In the edge type, light sources are placed at the edges of the backlight, so it is typically thinner than the direct type. Therefore, mobile devices, Personal Computers and little size LCD TVs use edge-type LED backlight nowadays. However, the conventional structure of Light Guide Plate in backlight module is a whole unit and this kind of structure will cause two tough problems. One is the complex coeffieients calculation of backlight block spread function (BSF) when determining suitable LED string brightness, because the LED strings of different blocks have influence on the brightness of other blocks. The other problem is the Clipping Artifact. After the modified of the LED strings brightness and the LCD Compensation, the real displayed image and input image are exactly not the same because the luminance of displayed image cannot exceed backlight luminance. This kind of distortion is called Clipping Artifact. To solve the first problem, we propose a novel structure of Light Guide Plate (LGP) consisting of multi-blocks in edge type backlight. The brightness of each LGP block is just controlled by its associated LED strings. Therefore, it avoids the complex coefficients calculation in consideration of the effect from other blocks. In addition, to reduce the energy consumption, the brightness of each LED string is dimmed based on the displayed image of each block. To the second problem, we propose a Clipping Ratio Threshold (CRT) method, which modifies the dimming brightness of LED string in each LGP block, to reduce the image distortion due to Clipping Artifact. With our method, the users of mobile devices can set the parameter "Clipping Threshold (CT)" by themselves according to their needs and achieve a good user experience.Finally, to achieve better display quality in large-size LCD TVs, direct type is first choice, because the edge type can’t get enough light luminance in the middle of display devices. However, the number of the LED units will be very large in direct type large-size TVs. Therefore, if controlling the LED unit point by point, the drive circuit will be complicated and suffer the expensive cost. In this paper, a ranks decomposition control method is proposed to reduce the LED backlight source drivers. The ranks decomposition control method decomposes the LED brightness matrix to row and column Pulse Width Modulation (PWM) control signals, driving the LED unit point by point. And this method archives adaptive dimming effects, such as high contrast ratio and reduced power consumption. To implement the matrix decomposition, we propose a Sub-matrix linear optimization algorithm to get the optimized row and column PWM control signals.