LinePilot digitizer recovers curves from line plots with manual and automatic calibration
LinePilot Digitizer: Line-Plot Recovery with Manual and Automatic Calibration
Computer Vision and Pattern RecognitionArtificial Intelligence
Summary
Getting numbers out of line graphs is tricky because you need to know exactly where the axes are and carefully extract the curves. The authors created LinePilot Digitizer, a tool that can recover these curves using color and three ways to calibrate the axes, including one that uses OCR (text recognition). They also built a benchmark called DigitizerBench to test how well different tools do at this task. LinePilot showed strong performance on this benchmark, balancing automatic processing and user control to improve accuracy.
What this means in practice
- •For data analysts: Extract numerical data accurately from line plots in scanned or digital images for further analysis using a tool with flexible calibration modes.
- •For document digitization teams: Automate recovery of data series from charts in legacy publications while allowing manual calibration to improve quality.
Authors
Fengbo Ma, Rayan Akhtar, Aakash H. Joshi, Xiaoting Li, Haijian Sun, Zhen Xiang, Xianyan Chen, Yiping Zhao
Abstract
Recovering numerical series from line plots requires accurate axis calibration and reliable curve extraction. We present LinePilot Digitizer (LinePilot), which combines continuous color-based curve recovery with three calibration modes: LinePilot (standard), LinePilot (enhanced), and LinePilot (OCR). We also introduce DigitizerBench, the first dedicated benchmark for systematically evaluating digitizer performance, using an orthogonal design spanning signal, rendering, and plot-structure factors with complementary automatic and human-guided evaluations. We evaluate performance using failure-penalized capped normalized root-mean-square error (FPC-NRMSE), which assigns unit loss to missing, unusable, or catastrophically inaccurate outputs. On DigitizerBench-Full, LinePilot (OCR) achieves the lowest mean FPC-NRMSE (0.672) and highest trusted usability (38.2%) among the tested automatic pipelines. On DigitizerBench-Lite, LinePilot (enhanced) achieves the lowest mean FPC-NRMSE (0.081), 100% output success, and highest trusted usability (93.3%). The orthogonal benchmark design further enables factor analysis to identify the factors that most significantly affect digitizer performance. Together, the three calibration modes provide a practical trade-off between automation, user control, and accuracy within a shared curve-recovery workflow.