Autonomous Apple Fruitlet Sizing with Next Best View Planning

TL;DR

This paper introduces a novel autonomous planning method for accurately sizing small apple fruitlets using strategic viewpoint selection and a dual-map system, outperforming existing approaches in accuracy and efficiency.

Contribution

The paper presents a new next-best-view planning approach that effectively sizes small fruitlets by combining semantic region sampling, attention-guided information gain, and a dual-map environment representation.

Findings

Improved sizing accuracy over state-of-the-art methods

Effective planning around small, densely packed fruitlets

Validated results on real robotic field data

Abstract

In this paper, we present a next-best-view planning approach to autonomously size apple fruitlets. State-of-the-art viewpoint planners in agriculture are designed to size large and more sparsely populated fruit. They rely on lower resolution maps and sizing methods that do not generalize to smaller fruit sizes. To overcome these limitations, our method combines viewpoint sampling around semantically labeled regions of interest, along with an attention-guided information gain mechanism to more strategically select viewpoints that target the small fruits' volume. Additionally, we integrate a dual-map representation of the environment that is able to both speed up expensive ray casting operations and maintain the high occupancy resolution required to informatively plan around the fruit. When sizing, a robust estimation and graph clustering approach is introduced to associate fruit detections across images. Through simulated experiments, we demonstrate that our viewpoint planner improves sizing accuracy compared to state of the art and ablations. We also provide quantitative results on data collected by a real robotic system in the field.