Artifact reduction method and apparatus and image processing method and apparatus

An image processing method, comprising:

• scaling down a resolution of each of at least one previous original image frame from a first resolution to a second resolution to obtain at least one previous low-resolution image frame and storing the at least one previous low-resolution image frame in a buffer unit;
• scaling down a resolution of a current original image frame from the first resolution to the second resolution to obtain a current low-resolution image frame, wherein the current original image frame and the current low-resolution image frame comprise a first plurality of pixels and a second plurality of pixels, respectively;
• generating a plurality of low-resolution weights according to a relation of the current low-resolution image frame and the at least one previous low-resolution image frame retrieved from the buffer unit, wherein the low-resolution weights are respectively corresponding to the second plurality of pixels of the current low-resolution image frame, and the step of generating the low-resolution weights comprises:
• for each pixel in the second plurality of pixels of the current low-resolution image frame, respectively executing following steps:
• (i) comparing a color or brightness value of a pixel in the current low-resolution image frame with a color or brightness value of the pixel in the at least one previous low-resolution image frame;
• and
• (ii) if the comparison result satisfies a predetermined condition, adjusting a color or brightness weight corresponding to the pixel in the low-resolution weights of the at least one previous low-resolution image frame to generate the color or brightness weight corresponding to the pixel in the low-resolution weights of the current low-resolution image frame;
• scaling up a number of the low-resolution weights to generate a plurality of high-resolution weights, wherein the high-resolution weights are respectively corresponding to the first plurality of pixels of the current original image frame;
• and
• performing an image processing procedure to the current original image frame by using the high-resolution weights to obtain an adjusted image frame, wherein the step of performing the image processing procedure to the current original image frame comprises:
• for each pixel in the first plurality of pixels of the current original image frame, respectively executing following steps:
• calculating a color or brightness value of a pixel in an adjusted image frame based on a color or brightness value of the pixel in the current original image frame, a color or brightness value of the pixel in a previous adjusted image frame, and a color or brightness weight corresponding to the pixel in the high-resolution weights.

An artifact reduction apparatus, comprising:

• a scaling-down circuitry scaling down a resolution of each of at least one previous original image frame from a first resolution to a second resolution to obtain at least one previous low-resolution image frame, and scaling down a resolution of a current original image frame from the first resolution to the second resolution to obtain and output a current low-resolution image frame, wherein the current original image frame and the current low-resolution image frame comprise a first plurality of pixels and a second plurality of pixels, respectively;
• a buffer unit, coupled to the scaling-down circuitry, receiving and storing the at least one previous low-resolution image frame and the current low-resolution image frame;
• an artifact detection circuitry coupled to the scaling-down circuitry and the buffer unit, wherein the artifact detection circuitry performs an artifact detection to the current low-resolution image frame to obtain and output a plurality of low-resolution weights in accordance with a relation of the current low-resolution image frame provided by the scaling-down circuitry and the at least one previous low-resolution image frame provided by the buffer unit, wherein the low-resolution weights are respectively corresponding to the second plurality of pixels in the current low-resolution image frame, and for each pixel in the second plurality of pixels of the current low-resolution image frame, the artifact detection circuitry respectively executes following steps:
• (i) comparing a color or brightness value of a pixel in the current low-resolution image frame with a color or brightness value of the pixel in the at least one previous low-resolution image frame
• (ii) if the comparison result satisfies a predetermined condition, adjusting a color or brightness weight corresponding to the pixel in the low-resolution weights of the at least one previous low-resolution image frame to generate the color or brightness weight corresponding to the pixel in the low-resolution weights of the current low-resolution image frame;
• a scaling-up circuitry coupled to the artifact detection circuitry, wherein the scaling-up circuitry scales up a number of the low-resolution weights to obtain and output a plurality of high-resolution weights, wherein the high-resolution weights are respectively corresponding to the first plurality of pixels in the current original image frame;
• and
• a filter circuitry coupled to the scaling-up circuitry, wherein the filter circuitry performs an image processing procedure to the current original image frame by using the high-resolution weights for reducing artifact of the current original image frame and obtaining an adjusted image frame,
• wherein for each pixel in the first plurality of pixels of the current original image frame, the filter circuitry respectively executes following steps:
• calculating a color or brightness value of a pixel in an adjusted image frame based on a color or brightness value of the pixel in the current original image frame, a color or brightness value of the pixel in a previous adjusted image frame, and a color or brightness weight corresponding to the pixel in the high-resolution weights.

An image processing apparatus, comprising:

• a scaling-down circuitry scaling down a resolution of each of at least one previous original image frame from a first resolution to a second resolution to obtain at least one previous low-resolution image frame, and scaling down a resolution of a current original image frame from the first resolution to the second resolution to obtain a current low-resolution image frame, wherein the current original image frame and the current low-resolution image frame comprise a first plurality of pixels and a second plurality of pixels, respectively;
• a buffer unit, coupled to the scaling-down circuitry, receiving and storing the at lest one previous low-resolution image frame and the current low-resolution image frame;
• a weight generation circuitry, coupled to the scaling-down circuitry and the buffer unit, generating a plurality of low-resolution weights according to a relation of the current low-resolution image frame provided by the scaling-down circuitry and the previous low-resolution image frame provided by the buffer unit, wherein the low-resolution weights are respectively corresponding to the second plurality of pixels in the current low-resolution image frame, and for each pixel in the second plurality of pixels of the current low-resolution image frame, the weight generation circuitry respectively executes following steps:
• (i) comparing a color or brightness value of a pixel in the current low-resolution image frame with a color or brightness value of the pixel in the at least one previous low-resolution image frame
• (ii) if the comparison result satisfies a predetermined condition, adjusting a color or brightness weight corresponding to the pixel in the low-resolution weights of the at least one previous low-resolution image frame to generate the color or brightness weight corresponding to the pixel in the low-resolution weights of the current low-resolution image frame;
• a scaling-up circuitry scaling up a number of the low-resolution weights to generate a plurality of high-resolution weights, wherein the high-resolution weights are respectively corresponding to the first plurality of pixels in the current original image frame;
• and
• an adjusting circuitry performing an image processing procedure to the current original image frame by using the high-resolution weights to obtain an adjusted image frame,
• wherein for each pixel in the first plurality of pixels of the current original image frame, the adjusting circuitry respectively executes following steps:
• calculating a color or brightness value of a pixel in an adjusted image frame based on a color or brightness value of the pixel in the current original image frame, a color or brightness value of the pixel in a previous adjusted image frame, and a color or brightness weight corresponding to the pixel in the high-resolution weights.