Content-Aware Raster to Vector Tracing with Edge Snapping

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional methods for converting raster images to vector drawings are inefficient and user-unfriendly, requiring manual tracing or batch processing, which leads to inaccurate results and excessive computational resources.

Innovation Solution

An image conversion system that performs content-aware conversion by detecting edges in raster images, allowing users to trace edges with precision and accuracy using a graphical user interface, and generating smooth parametric curves like Bézier curves based on edge points, even if the tracing path does not exactly follow the edge.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If manual tracing is used to convert raster images to vector drawings, then user control over the conversion process is improved, but the time required and user effort increase significantly

Engineering Contradiction:
Improveuser controlVSAvoidconversion time
Core Design Contradiction:
Ease of operationVSLoss of time

Solution Approach 1:

The system performs preliminary edge detection on the raster image to pre-identify all edge locations before the user begins tracing. This pre-processing step creates a foundation that accelerates the subsequent tracing operation, allowing the system to instantly recognize when the cursor is near an edge and automatically snap to it, rather than requiring the user to manually trace every pixel boundary.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system replaces the mechanical manual tracing process with an automated edge detection and snapping mechanism. Instead of relying on the user's manual cursor movement to define edges, the system uses image processing algorithms to detect edges and automatically guides the cursor to follow them, substituting computational automation for manual mechanical tracing while preserving user oversight.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Productivity

If batch conversion is used to convert entire raster images to vector drawings, then processing speed is improved, but accuracy and relevance of the resulting vector curves decrease

Engineering Contradiction:
Improveprocessing speedVSAvoidvector curve accuracy
Core Design Contradiction:
ProductivityVSManufacturing precision

Solution Approach 1:

The system applies local quality by performing edge detection and vector conversion only in the local region where the user is currently tracing, rather than processing the entire image uniformly. This allows the system to maintain high accuracy in the active tracing area while avoiding the generation of unnecessary vector curves in other regions, thereby improving both precision and efficiency.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The system performs partial action by converting only the portion of the raster image that the user is actively tracing at any given moment, rather than converting the entire image at once. This selective approach generates fewer unnecessary vector curves and allows the user to control which edges are converted, improving accuracy while maintaining processing efficiency.

Inventive Principle:
Principle #16Partial or excessive action

3Ease of operation

If snapping pen tool is used to assist manual tracing, then ease of tracing is improved, but the resulting paths become jagged and fail to capture overall edge shapes

Engineering Contradiction:
Improvetracing easeVSAvoidpath smoothness
Core Design Contradiction:
Ease of operationVSShape

Solution Approach 1:

The system applies curvature by fitting smooth parametric curves (such as Bézier curves) to the detected edge points during the tracing process. Instead of creating jagged paths from discrete snapping points, the system continuously generates smooth curves that approximate the underlying edge geometry, preserving the overall shape and visual quality of the traced objects.

Inventive Principle:
Principle #14Spheroidality (Curvature)

Data Source

PatentUS10930033B1Content aware tracing for generating vector drawings
Publication Date: 2021.02.23 ADOBE INC
  • US10930033B1 patent drawing
  • US10930033B1 patent drawing
  • US10930033B1 patent drawing

AI summary

Techniques and systems are provided for generating a vector drawing from a raster image. For example, an edge map for a raster image can be obtained, and a point of the edge map can be determined that corresponds to an input position associated with user interaction with a graphical user interface. A search process can be performed to determine an edge point of the edge map corresponding to the input position. Using the search process, the edge point is determined from a neighborhood of points associated with the point of the edge map. Using a plurality of edge points of the edge map (including the determined edge point), a vector drawing segment for a vector drawing representing the raster image can be generated.