Segmentation Maps for Selecting Overlapping Objects in Digital Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional techniques struggle with identifying and selecting objects in digital images, especially when they lack predefined separation, leading to user frustration and inefficient resource consumption, particularly in scenarios with complex shapes or overlapping objects.
Innovation Solution
A digital image processing system employs instance segmentation, structure identification, element identification, and object association techniques using machine learning models to identify and bound objects, providing visual guides for selection and manipulation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple objects are provided in a digital image without predefined separation, then the digital image can include complex layouts with overlaps and irregular spacing, but difficulty in selection of these objects increases
Solution Approach 1:
The patent applies segmentation by dividing the digital image into multiple segments or regions, each potentially containing one or more objects. By segmenting the image space and assigning segment identifiers to pixels, the system creates discrete selectable units even when objects overlap or lack predefined separation, thereby maintaining layout flexibility while enabling object selection.
Solution Approach 2:
The patent introduces an intermediary data structure (segmentation map or segment identifier field) that mediates between the complex visual layout and the selection mechanism. This intermediary layer assigns unique segment identifiers to different regions, allowing objects to be selected based on their segment membership rather than relying on predefined object boundaries, thus resolving the selection difficulty.
2Device complexity
If conventional techniques are used for object selection in complex digital images, then implementation is simple, but user frustration increases and computational resources are consumed inefficiently
Solution Approach 1:
The patent applies preliminary action by performing image segmentation and assigning segment identifiers to pixels before the selection operation occurs. This pre-processing step creates an organized structure that enables efficient object selection and manipulation, avoiding the need for complex real-time analysis during user interaction, thereby improving productivity while maintaining reasonable system complexity.
3Shape
If objects have complex shapes or overlaps, then the digital image can represent realistic scenarios, but identification and selection of individual objects becomes difficult
Solution Approach 1:
The patent applies dimensionality change by introducing a new dimension (segment identifier dimension) to the image data. Instead of relying solely on spatial coordinates and shape characteristics in the original image plane, the system adds segment identifier labels that provide an additional dimension for object identification and selection, making it easier to distinguish and manipulate objects with complex shapes or overlaps.
Data Source
AI summary
Object identification techniques from a digital image are described. In an implementation, edges of an object are determined by analyzing gradients from a digital image. A structure of the object is computed by detecting line segments from the digital image. A boundary of the object is defined based on the edges and the structure. A display of the object is edited in a user interface based on the boundary using an edit operation.