Encoding of pre-processed image frames

Patent metadata

Field	Value
Publication number	US-12563238-B2
Application number	US-202418410396-A
Country	US
Kind code	B2
Filing date	Jan 11, 2024
Priority date	Feb 9, 2023
Publication date	Feb 24, 2026
Grant date	Feb 24, 2026

Abstract

A method for encoding an image frame, performed by an image processing device, comprising obtaining image data, and identifying an image area in an image frame based on that the image area fulfilling an identification criterion. The method further comprises determining a bit depth reduction factor for the identified image area by analyzing the image data in the identified image area, and replacing some of the bit values of the pixel values in the identified image area with dummy values. How many of the bit values that are replaced with dummy values is defined by the bit depth reduction factor. The method comprises encoding the image frame upon said some of the bit values having been replaced in the identified image area.

First claim

The invention claimed is:

1 . An image processing device for encoding an image frame, the image processing device comprising processing circuitry being configured to cause the image processing device to:
obtain image data, in terms of pixel values, representing the image frame, wherein each pixel value is represented by respective bit values ranging from a most significant bit value to a least significant bit value; identify an image area in the image frame based on that the image area fulfilling an identification criterion; determine a bit depth reduction factor for the identified image area by analyzing the image data in the identified image area; replace some of the bit values of the pixel values in the identified image area with dummy values, starting from the least significant bit value and continuing towards the most significant bit value, wherein how many of the bit values that are replaced with dummy values is defined by the bit depth reduction factor; and encode the image frame upon said some of the bit values having been replaced in the identified image area, wherein the image area is encoded to an independent block, a predictive block, or a bidirectional predictive block, and wherein the cost for encoding the image area to the independent block is increased from a nominal value when the identified image area comprises dummy values.

2 . The image processing device according to claim 1 , wherein the identification criterion is based on one or more of:
signal to noise ratio calculation of the image area, background-foreground classification of the image area, region of interest classification of the image area, object classification of an object depicted in the image area, or object identification of an object depicted in the image area.

3 . The image processing device according to claim 1 , wherein the identification criterion is fulfilled when one or more of the following is true:
signal to noise ratio for the image area is lower than a threshold value, the image area depicts a background in the image frame, the image area depicts a non-region of interest in the image frame, or the image area depicts an object that is outside an object class of interest.

4 . The image processing device according to claim 1 , wherein analyzing the image data in the identified image area comprises calculating a signal to noise ratio for the image area, and wherein the bit depth reduction factor is a function of the calculated signal to noise ratio for the image area.

5 . The image processing device according to claim 1 , wherein analyzing the image data in the identified image area comprises performing object classification to identify objects in the identified image area, and wherein the bit depth reduction factor is a function of which type of object was identified in the identified image area.

6 . The image processing device according to claim 1 , wherein analyzing the image data in the identified image area comprises checking whether any metadata is available for the identified image area or not, and wherein the bit depth reduction factor is set higher if any metadata is available than if any metadata is not available.

7 . The image processing device according to claim 1 , wherein analyzing the image data in the identified image area comprises comparing a distance to any object depicted in the identified image area with a distance interval, and wherein the bit depth reduction factor is higher if the distance to said any object is outside the distance interval than if the distance to said any object is inside the distance interval.

8 . The image processing device according to claim 1 , wherein the processing circuitry further is configured to cause the image processing device to, as part of replacing said some of the bit values:
decrease bit depth of the pixel values in the identified image area from a first value to a second, different, value, by removing as many bit values from the pixel values in the identified image area as given by the bit depth reduction factor, starting from the least significant bit value and continuing towards the most significant bit value; and restore the bit depth of the pixel values in the identified image area back to the first value by adding the dummy values where the bit values were removed.

9 . The image processing device according to claim 1 , wherein each pixel value has a luminance part and two chrominance parts, and wherein one respective bit depth reduction factor is determined for each of the luminance part and the two chrominance parts.

10 . The image processing device according to claim 1 , wherein the dummy values per bit value are equal for all pixels in the image area.

11 . The image processing device according to claim 1 , wherein during the encoding, the identified image area is flagged as comprising dummy values.

12 . A method for encoding an image frame, the method being performed by an image processing device, the method comprising:
obtaining image data, in terms of pixel values, representing the image frame, wherein each pixel value is represented by respective bit values ranging from a most significant bit value to a least significant bit value; identifying an image area in the image frame based on that the image area fulfils an identification criterion; determining a bit depth reduction factor for the identified image area by analyzing the image data in the identified image area; replacing some of the bit values of the pixel values in the identified image area with dummy values, starting from the least significant bit value and continuing towards the most significant bit value, wherein how many of the bit values that are replaced with dummy values is defined by the bit depth reduction factor; and encoding the image frame upon said some of the bit values having been replaced in the identified image area, wherein the image area is encoded to an independent block, a predictive block, or a bidirectional predictive block, and wherein the cost for encoding the image area to the independent block is increased from a nominal value when the identified image area comprises dummy values.

13 . A non-transitory computer readable recording medium comprising a computer program for encoding an image frame, the computer program comprising computer code which, when run on processing circuitry of an image processing device, causes the image processing device to:
obtain image data, in terms of pixel values, representing the image frame, wherein each pixel value is represented by respective bit values ranging from a most significant bit value to a least significant bit value; identify an image area in the image frame based on that the image area fulfils an identification criterion; determine a bit depth reduction factor for the identified image area by analyzing the image data in the identified image area; replace some of the bit values of the pixel values in the identified image area with dummy values, starting from the least significant bit value and continuing towards the most significant bit value, wherein how many of the bit values that are replaced with dummy values is defined by the bit depth reduction factor; and encode the image frame upon said some of the bit values having been replaced in the identified image area, wherein the image area is encoded to an independent block, a predictive block, or a bidirectional predictive block, and wherein the cost for encoding the image area to the independent block is increased from a nominal value when the identified image area comprises dummy values.

Assignees

• Axis Ab

Inventors

• Edpalm Viktor
• Yuan Song

Classifications

• H04N19/20

  using video object coding · CPC title

• H04N19/186

  the unit being a colour or a chrominance component · CPC title

• H04N19/184

  the unit being bits, e.g. of the compressed video stream · CPC title

• H04N19/182

  the unit being a pixel · CPC title

• H04N19/176

  the region being a block, e.g. a macroblock · CPC title