Segmentation-guided Layer-wise Image Vectorization with Gradient Fills

TL;DR

This paper introduces a segmentation-guided, layer-wise vectorization method that converts raster images into vector graphics with gradient fills, improving visual quality and topology without dataset dependence.

Contribution

The proposed framework uniquely integrates gradient-aware segmentation and a differentiable renderer for dataset-independent raster-to-vector conversion with gradient fills.

Findings

Produces vector graphics with improved visual quality.

Achieves better layer-wise topology compared to prior methods.

Operates as a model-free, dataset-independent tool.

Abstract

The widespread use of vector graphics creates a significant demand for vectorization methods. While recent learning-based techniques have shown their capability to create vector images of clear topology, filling these primitives with gradients remains a challenge. In this paper, we propose a segmentation-guided vectorization framework to convert raster images into concise vector graphics with radial gradient fills. With the guidance of an embedded gradient-aware segmentation subroutine, our approach progressively appends gradient-filled B\'ezier paths to the output, where primitive parameters are initiated with our newly designed initialization technique and are optimized to minimize our novel loss function. We build our method on a differentiable renderer with traditional segmentation algorithms to develop it as a model-free tool for raster-to-vector conversion. It is tested on various inputs to demonstrate its feasibility, independent of datasets, to synthesize vector graphics with improved visual quality and layer-wise topology compared to prior work.