LaneCPP: Continuous 3D Lane Detection using Physical Priors

TL;DR

LaneCPP addresses monocular 3D lane detection by introducing a continuous 3D lane representation based on $3$D B-Splines and integrating physics-inspired priors through a regularization framework. It couples a geometry-aware spatial transformation to lift 2D image features into a learned 3D road-space feature map, guided by surface hypotheses and a surface-height loss. The method achieves state-of-the-art F1-scores and reduced $X$- and $Z$-errors on the OpenLane benchmark and demonstrates competitive results on Apollo 3D Synthetic, validating the benefit of combining analytic geometric priors with learnable 3D feature representations. Overall, LaneCPP shows that incorporating priors and geometry-aware processing yields robust, predictable 3D lane perception suitable for autonomous driving scenarios.

Abstract

Monocular 3D lane detection has become a fundamental problem in the context of autonomous driving, which comprises the tasks of finding the road surface and locating lane markings. One major challenge lies in a flexible but robust line representation capable of modeling complex lane structures, while still avoiding unpredictable behavior. While previous methods rely on fully data-driven approaches, we instead introduce a novel approach LaneCPP that uses a continuous 3D lane detection model leveraging physical prior knowledge about the lane structure and road geometry. While our sophisticated lane model is capable of modeling complex road structures, it also shows robust behavior since physical constraints are incorporated by means of a regularization scheme that can be analytically applied to our parametric representation. Moreover, we incorporate prior knowledge about the road geometry into the 3D feature space by modeling geometry-aware spatial features, guiding the network to learn an internal road surface representation. In our experiments, we show the benefits of our contributions and prove the meaningfulness of using priors to make 3D lane detection more robust. The results show that LaneCPP achieves state-of-the-art performance in terms of F-Score and geometric errors.

Figures (15)

• Figure 1: Our approach: First, front-view image $I$ is propagated through the backbone extracting multi-scale feature maps. These are transformed to 3D using our spatial transformation and then fused to obtain a single 3D feature map. Feature pooling is applied to obtain features for each line proposal that are propagated through fully connected layers to obtain the parameters for our line representation. Finally, prior knowledge is exploited to regularize the lane representation and to produce surface hypotheses for the spatial transformation.
• Figure 2: Our 3D lane line representation: For each proposal $\boldsymbol{\bar{f}}$ (purple lines), line geometry is described by 3D B-Splines with control points $\boldsymbol{c}_k$ (green dots). Each control point is determined by the offsets $\alpha_k, \, \beta_k$ from the control points of the initial proposal in normal direction (orange vectors). Additionally, visibility $v(t)$ is modeled by splines with 1D control points $\gamma_k$.
• Figure 3: Illustration of different priors expressed by line tangents and surface normals.
• Figure 4: Our proposed spatial transformation module. First, several road surface hypotheses are defined (a) to which front-view features are lifted (b) and weighted according to the predicted depth distribution. Afterwards, point features are aggregated in a weighted manner to obtain the 3D feature map (c).
• Figure 5: Qualitative comparison of our model trained with prior regularization to the same model without regularization both trained on OpenLane300 with main differences highlighted by arrows. As a reference ground truth lines are visualized dashed.