US 12,755,358 B2Granted

HDR tone mapping system and method with semantic segmentation

AssigneeMediatek Inc.

InventorsHuei-Han Jhuang, Jan-Wei Wang, Po-Yu Huang, Ying-Jui Chen, Chi-Cheng Ju

Life of this patent

Estimating the term…

Loading patent…

Abstract

A HDR tone mapping system includes several modules. A semantic segmentation module is used to extract semantic information from the input image. An image decomposition module is used to decompose the input image to a high-bit base layer and a detail layer. A statistics module is used to generate statistics of pixels of the input image according to the semantic information. A curve computation module is used to generate a tone curve from the statistics. A compression module is used to compress the high-bit base layer to a low-bit base layer according to the tone curve, the statistics and the semantic information. A detail adjustment module is used to tune the detail layer according to the semantic information and the statistics to generate an adjusted detail layer. An image reconstruction module is used to combine the adjusted detail layer and the low-bit base layer to generate an output image.

Claims18

Dependent claims

A HDR (High Dynamic Range) tone mapping system comprising one or more processors configured to:

receive an input image and extract semantic information from the input image by performing semantic segmentation on image content of the input image to identify a plurality of semantic objects in the input image;
decompose the input image to a base layer having a first bit depth and a detail layer according to the semantic information;
generate statistics of pixels of the input image according to the semantic information;
generate a tone curve according to the statistics of the pixels;
compress the base layer having the first bit depth to the base layer having a second bit depth less than the first bit depth, wherein, for each semantic object of the plurality of semantic objects, pixels in the base layer having the first bit depth that belong to the semantic object are compressed together according to a respective tone curve, respective statistics, and respective semantic information corresponding to the semantic object;
tune the detail layer according to the semantic information and the statistics to generate an adjusted detail layer;
and
combine the adjusted detail layer and the base layer having the second bit depth to generate an output image.

5 more elements

A HDR (High Dynamic Range) tone mapping method implemented by a computer, the method comprising:

receiving an input image and extracting semantic information from the input image by performing semantic segmentation on image content of the input image to identify a plurality of semantic objects in the input image;
decomposing the input image to a base layer having a first bit depth and a detail layer according to the semantic information;
generating statistics of pixels of the input image according to the semantic information;
generating a tone curve according to the statistics of the pixels;
compressing the base layer having the first bit depth to a base layer having a second bit depth less than the first bit depth, wherein, for each semantic object of the plurality of semantic objects, pixels in the base layer having the first bit depth that belong to the semantic object are compressed together according to a respective tone curve, respective statistics, and respective semantic information corresponding to the semantic object;
tuning the detail layer according to the semantic information and the statistics to generate an adjusted detail layer;
and
combining the adjusted detail layer and the base layer having the second bit depth to generate an output image.

5 more elements

Dependencies are inferred from the claim text.