Player Mask Generation for Depth-Color Video Integration

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current image processing technologies for capturing live player data in gaming and entertainment environments face challenges such as noise, artifacts, and temporal and spatial discontinuities, leading to reduced image quality when combining depth information and color data from a capture device with video data.

Innovation Solution

The implementation of systems and methods that apply noise reduction filters, adaptive statistical color background models, edge blending, and depth mask smoothing to enhance the quality of the player image, using techniques like bicubic down/upsampling and thresholding to improve the accuracy and smoothness of the player mask, ensuring seamless integration with the background video.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If depth information and color data are captured and combined to depict the live player in the video scene, then motion capture realism is improved, but noise and artifacts are introduced reducing image quality

Engineering Contradiction:
Improvemotion capture realismVSAvoidnoise and artifacts
Core Design Contradiction:
ReliabilityVSObject-affected harmful factors

Solution Approach 1:

A player mask is introduced as an intermediary element to separate the player from the background. The mask is generated from depth information and used to selectively apply processing operations, allowing the player image to be extracted and combined with video data while reducing noise and artifacts through controlled blending operations.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

Different processing operations are applied to different regions of the image based on the player mask. The mask enables local quality enhancement by applying noise reduction and blending operations specifically to player regions while preserving background integrity, rather than processing the entire image uniformly.

Inventive Principle:
Principle #3Local quality

2Measurement precision

If the player mask is generated from depth information, then player segmentation is achieved, but spatial discontinuities and rough edges occur reducing image quality

Engineering Contradiction:
Improveplayer segmentation accuracyVSAvoidedge smoothness
Core Design Contradiction:
Measurement precisionVSShape

Solution Approach 1:

The player mask undergoes dynamic processing including blurring operations that adapt to the mask's characteristics. The mask is blurred to smooth edges while maintaining player segmentation accuracy, creating a dynamic balance between precise segmentation and smooth edge appearance in the final composited image.

Inventive Principle:
Principle #15Dynamics

3Object-affected harmful factors

If noise reduction filters are applied to the color data, then image quality is improved, but processing time and computational complexity increase

Engineering Contradiction:
Improvenoise reductionVSAvoidprocessing time
Core Design Contradiction:
Object-affected harmful factorsVSLoss of time

Solution Approach 1:

The image processing is segmented into distinct regions using the player mask. Noise reduction filters are applied selectively only to player regions rather than the entire image, reducing overall processing time and computational complexity while maintaining image quality in critical areas.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentUS8823745B2Image processing based on depth information and color data of a scene
Publication Date: 2014.09.02 NOO INC
  • US8823745B2 patent drawing
  • US8823745B2 patent drawing
  • US8823745B2 patent drawing

AI summary

Systems, apparatuses, methods, and computer program products perform image processing in an environment in which depth information and color data of a scene including a player and a background are received from a capture device, and in which an image of the player combined with video data is output.