Image Blending Using Depth-Aware Segmentation for Accurate Field Integration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Dual-camera modules in terminal devices cannot effectively blend a target object into a corresponding field depth within an input image based on the object's depth information, limiting the integration of virtual reality elements.
Innovation Solution
An image processing apparatus comprising an image capture device, filter, depth estimation unit, and mixture unit that captures an original image, generates depth maps, and blends the target object into a predetermined field depth of the input image using a Z-test method, allowing for accurate depth integration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the terminal device blends the target object into the input image without depth information integration, then the blending process is simple, but the target object cannot be accurately positioned in the corresponding field depth
Solution Approach 1:
The patent segments the image processing into distinct modules: depth map generation module, object detection module, and blending module. Each module processes specific information independently (depth maps from dual cameras, object detection from extracted images, blending from processed data), then integrates results. This segmentation enables accurate depth positioning while managing complexity through modular architecture.
Solution Approach 2:
The patent introduces a depth dimension to the traditional 2D image blending process. By generating depth maps from dual camera images and integrating depth information with object detection results, the system transitions from flat 2D blending to 3D-aware blending. This allows target objects to be positioned at correct depth planes within the input image, achieving accurate field depth integration.
2Adaptability or versatility
If the terminal device uses dual-cameras module for image processing, then the application scope is expanded, but the integration of virtual reality elements is limited without proper depth blending
Solution Approach 1:
The patent creates a universal image processing system that handles multiple functions: capturing images with dual cameras, generating depth maps, detecting objects, and blending results. The same processed image data can be applied to various scenarios including virtual reality integration, image editing, and enhanced display. This multi-functional approach expands application scope while maintaining reliable blending through consistent depth-aware processing.
Solution Approach 2:
The patent implements feedback mechanisms where depth map information is continuously refined and used to guide object detection and blending operations. The system uses depth information from dual cameras to verify and adjust object positioning, creating a feedback loop that ensures accurate integration. This feedback process enhances reliability by continuously validating blending results against depth constraints.
3Ease of operation
If the terminal device covers only the target object on the input image without depth integration, then the processing is fast, but the visual realism and user experience are reduced
Solution Approach 1:
The patent performs preliminary actions by pre-processing dual camera images to generate depth maps before the actual blending operation. Object detection and feature extraction are performed in advance on the captured images. These preliminary computations prepare all necessary depth and object information beforehand, so that when blending occurs, the system can quickly integrate results without real-time computation delays, thus improving user experience while controlling processing time.
Data Source
AI summary
An image process apparatus includes an image capture device, a filter, a depth estimation unit, and a mixture unit. The image capture device captures an original image including at least one target object and generates a first depth map corresponding to the original image. The filter selects the at least one target object from the original image according to the first depth map and generates a temporary image including the at least one target object. The temporary image has a depth information of the at least one target object. The depth estimation unit generates a second depth map corresponding to an input image according to the input image. The mixture unit blends the temporary image into a predetermined field depth of the input image to generate a blending image including the input image and the at least one target object according to the depth information and the second depth map.