Abstract: A training method for video stabilization and an image processing device using the same are proposed. The method includes the following steps. An input video including low dynamic range (LDR) images is received. The LDR images are converted to high dynamic range (HDR) images by using a first neural network. A feature extraction process is performed to obtain features based on the LDR images and the HDR images. A second neural network for video stabilization is trained according to the LDR images and the HDR images based on a loss function by minimizing a loss value of the loss function to generate stabilized HDR images in a time-dependent manner, where the loss value of the loss function depends upon the features. An HDR classifier is constructed according to the LDR images and the HDR images. The stabilized HDR images are classified by using the HDR classifier to generate a reward value, where the loss value of the loss function further depends upon the reward value.

Abstract: A training method for video stabilization and an image processing device using the same are proposed. The method includes the following steps. An input video including low dynamic range (LDR) images is received. The LDR images are converted to high dynamic range (HDR) images by using a first neural network. A feature extraction process is performed to obtain features based on the LDR images and the HDR images. A second neural network for video stabilization is trained according to the LDR images and the HDR images based on a loss function by minimizing a loss value of the loss function to generate stabilized HDR images in a time-dependent manner, where the loss value of the loss function depends upon the features. An HDR classifier is constructed according to the LDR images and the HDR images. The stabilized HDR images are classified by using the HDR classifier to generate a reward value, where the loss value of the loss function further depends upon the reward value.

Abstract: A training method for video stabilization and an image processing device using the same are proposed. The method includes the following steps. An input video including low dynamic range (LDR) images is received. The LDR images are converted to high dynamic range (HDR) images by using a first neural network. A second neural network for video stabilization is trained to generate stabilized HDR images in a time-dependent manner.

Abstract: A training method for video stabilization and an image processing device using the same are proposed. The method includes the following steps. An input video including low dynamic range (LDR) images is received. The LDR images are converted to high dynamic range (HDR) images by using a first neural network. A second neural network for video stabilization is trained to generate stabilized HDR images in a time-dependent manner.

Abstract: An estimation method of a predicted motion vector for an image block having one or more pixels, which includes calculating a pixel difference value corresponding to a current frame and a reference frame for each pixel of the image block, determining a pixel difference area according to the pixel difference values corresponding to the plurality of pixels, and determining a predicted motion vector according to the pixel difference area.

Abstract: A motion estimation method is provided, which includes following steps: dividing a first frame to be estimated into a plurality of area units, in which each of the area units includes a plurality of blocks; and assigning a set of motion vector values to each of the area units, in which the set of motion vector values includes a plurality of predetermined motion vector values, and each of the predetermined motion vector values is assigned to at least one block in each of the area units.

Abstract: An image processing circuit and an image processing method are disclosed. The image processing circuit comprises a block matching unit, a multiplexer, an arbiter, and a motion compensation circuit. The block matching unit calculates an alternating current (AC) sum of absolute difference (SAD) and a direct current (DC) sum of absolute difference (SAD) according to a current block in a current image and a reference block in a reference image. The arbiter controls the multiplexer selectively to output the AC SAD or the DC SAD according to an arbitration rule related to a scene characteristic of the current image. The motion compensation circuit executes motion compensation according to the AC SAD or the DC SAD outputted by the multiplexer.

Abstract: A block matching method for estimating a motion vector of an estimation block of an image frame is provided, which includes comparing the estimation block with at least one reference block corresponding to a first object to obtain a plurality of pixel difference values, determining a mask area corresponding to the first object and a calculation area corresponding to a second object in the estimation block, and performing blocking matching operations on the calculation area to determine a motion vector of the second object as the motion vector of the estimation block.

Abstract: An estimation method of a predicted motion vector for an image block having one or more pixels, which includes calculating a pixel difference value corresponding to a current frame and a reference frame for each pixel of the image block, determining a pixel difference area according to the pixel difference values corresponding to the plurality of pixels, and determining a predicted motion vector according to the pixel difference area.

Abstract: An image processing circuit and an image processing method are disclosed. The image processing circuit comprises a block matching unit, a multiplexer, an arbiter, and a motion compensation circuit. The block matching unit calculates an alternating current (AC) sum of absolute difference (SAD) and a direct current (DC) sum of absolute difference (SAD) according to a current block in a current image and a reference block in a reference image. The arbiter controls the multiplexer selectively to output the AC SAD or the DC SAD according to an arbitration rule related to a scene characteristic of the current image. The motion compensation circuit executes motion compensation according to the AC SAD or the DC SAD outputted by the multiplexer.

Abstract: A motion estimation method is provided, which includes following steps: dividing a first frame to be estimated into a plurality of area units, in which each of the area units includes a plurality of blocks; and assigning a set of motion vector values to each of the area units, in which the set of motion vector values includes a plurality of predetermined motion vector values, and each of the predetermined motion vector values is assigned to at least one block in each of the area units.

Abstract: A image-driving method for a display includes receiving an image frame and registering at least a part of the image frame, wherein the image frame is divided into a prior-part frame and a post-part frame; respectively conducting a first luminance adjustment on the prior-part frame and the post-part frame so as to take the adjustment results as a first part of a first image frame and a first part of a second image frame; filling the previous received image frame after a second luminance processing into a second part of the first image frame; filling the presently received image frame after a second luminance processing into a second part of the second image frame; outputting the complete first image frame and the complete second image frame for successive displaying.

Abstract: A block matching method for estimating a motion vector of an estimation block of an image frame is provided, which includes comparing the estimation block with at least one reference block corresponding to a first object to obtain a plurality of pixel difference values, determining a mask area corresponding to the first object and a calculation area corresponding to a second object in the estimation block, and performing blocking matching operations on the calculation area to determine a motion vector of the second object as the motion vector of the estimation block.

Abstract: An impulse-type driving method for a liquid crystal display (LCD) is used for driving a pixel array of an LCD panel. The method includes providing a set of impulse control signals to a source driver. The source driver, according to the set of impulse control signals, drives the pixel array. The set of impulse control signals includes a command signal. The command signal includes a field of determining data voltage polarity and a command field. According to a time sequence, the field of determining data voltage polarity provides a polarity data for determining a voltage polarity output by the source driver. The command field and the field of determining data voltage polarity are consecutively and alternatively output, in which the command field allows to add a dynamic command in accordance with a desired action.

Abstract: An output enable signal transformation device for a gate driver in an LCD device includes a reception terminal coupled to a timing generator of the LCD device for receiving an enable synchronization signal, an enable clock signal and a plurality of enable control signals generated by the timing generator, a shift register module coupled to the reception terminal for shifting the enable synchronization signal according to the enable clock signal, a multiplexer module coupled to the shift register module and the timing generator for generating a plurality of output enable signals according to the enable synchronization signal and the plurality of enable control signals, and an output terminal coupled to the multiplexer module and a logic circuit of the gate driver for outputting the plurality of output enable signals to the logic circuit.

Abstract: A method for driving an LCD monitor includes receiving image data corresponding to a pixel of the LCD monitor, comparing pixel values of a first frame data and a second frame data in the image data, dividing the second frame data into a plurality of sub-frame data when a difference between the first frame data and the second frame is greater than a predetermined value, adjusting pixel values of the sub-frame data according to the pixel value of the second frame data, and sequentially displaying the sub-frame data by the pixel.

Abstract: An output enable signal transformation device for a gate driver in an LCD device includes a reception terminal coupled to a timing generator of the LCD device for receiving an enable synchronization signal, an enable clock signal and a plurality of enable control signals generated by the timing generator, a shift register module coupled to the reception terminal for shifting the enable synchronization signal according to the enable clock signal, a multiplexer module coupled to the shift register module and the timing generator for generating a plurality of output enable signals according to the enable synchronization signal and the plurality of enable control signals, and an output terminal coupled to the multiplexer module and a logic circuit of the gate driver for outputting the plurality of output enable signals to the logic circuit.

Abstract: An impulse-type driving method for a liquid crystal display (LCD) is used for driving a pixel array of an LCD panel. The method includes providing a set of impulse control signals to a source driver. The source driver, according to the set of impulse control signals, drives the pixel array. The set of impulse control signals includes a command signal. The command signal includes a field of determining data voltage polarity and a command field. According to a time sequence, the field of determining data voltage polarity provides a polarity data for determining a voltage polarity output by the source driver. The command field and the field of determining data voltage polarity are consecutively and alternatively output, in which the command field allows to add a dynamic command in accordance with a desired action.

Abstract: A image-driving method for a display includes receiving an image frame and registering at least a part of the image frame, wherein the image frame is divided into a prior-part frame and a post-part frame; respectively conducting a first luminance adjustment on the prior-part frame and the post-part frame so as to take the adjustment results as a first part of a first image frame and a first part of a second image frame; filling the previous received image frame after a second luminance processing into a second part of the first image frame; filling the presently received image frame after a second luminance processing into a second part of the second image frame; outputting the complete first image frame and the complete second image frame for successive displaying.

Abstract: A method for driving a flat display having multiple pixels is provided. Firstly, whether a display mode of the flat display is a normal mode or an impulse-driving mode is determined according to a control signal. Next, the pixels are driven according to a data signal and a first set of reference voltages when the display mode is the normal mode, such that the relationship between the gray level and the light intensity of the pixels is a first gamma curve. Afterwards, the pixels are driven according to the data signal and a second set of reference voltages when the display mode is the impulse-driving mode, such that the relationship between the gray level and the light intensity of the pixels is a second gamma curve, and a second normalization curve is close to a first normalization curve.