Autonomous Harvester Positioning Architecture With Monocular Vision
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Orchard harvesting technologies face challenges such as irregular tree spacing, varying tree sizes, complex canopy structures, low connectivity, and dusty environments, which hinder the integration of advanced automation technologies, leading to inefficient and labor-intensive harvesting methods.
Innovation Solution
An autonomous harvester equipped with advanced sensory and navigational technology, capable of generating a three-dimensional representation of the orchard environment, identifying tree emergence points, and implementing precise harvesting actions to minimize damage to both produce and trees.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If monocular images are used for tree localization, then device complexity is reduced, but measurement precision deteriorates due to inability to directly capture depth information
Solution Approach 1:
The patent replaces traditional stereo vision or LiDAR systems with a monocular camera system that uses computational geometry and virtual rays to achieve 3D localization. This substitutes complex mechanical sensing systems with a simpler optical system combined with algorithmic processing, reducing hardware complexity while maintaining measurement precision through mathematical modeling of the camera's pose and virtual ray projection.
Solution Approach 2:
The patent introduces virtual rays as an intermediary mathematical construct to bridge the gap between 2D monocular images and 3D tree positions. These virtual rays, combined with pose information, serve as a mediator that enables precise localization without requiring direct depth sensing hardware, thus resolving the contradiction between simplicity and precision.
2Productivity
If advanced automation technologies are integrated into orchard harvesting, then productivity is improved, but reliability deteriorates due to challenging environmental conditions including low connectivity and dust
Solution Approach 1:
The system employs self-contained monocular vision and pose estimation capabilities that operate autonomously without external connectivity. The autonomous harvester uses its own sensors and computational systems to navigate and harvest, making the system self-sufficient in low-connectivity orchard environments while maintaining high productivity through automation.
Solution Approach 2:
The patent adapts the system to orchard conditions by modifying operational parameters such as image capture frequency, pose estimation thresholds, and virtual ray projection methods. These parameter adjustments optimize performance for dusty environments and maintain reliability while preserving automation productivity.
3Manufacturing precision
If precise tree localization is implemented, then manufacturing precision is improved for harvesting actions, but device complexity increases due to sophisticated sensing and computational requirements
Solution Approach 1:
The patent replaces complex multi-sensor systems with a simplified monocular vision approach combined with computational geometry. By using virtual rays and pose information derived from standard cameras, the system achieves precise tree localization without requiring sophisticated hardware, thus improving manufacturing precision while controlling device complexity.
Solution Approach 2:
The patent transforms 2D monocular image data into 3D spatial understanding through virtual ray projection and pose estimation. This dimensional transformation enables precise harvesting actions by converting simple camera images into accurate tree position and orientation information, achieving precision without proportionally increasing hardware complexity.
Data Source
AI summary
An autonomous harvesting machine for orchard operating environments is described. The autonomous harvesting machine uses machine vision techniques to identify and triangulate features in the operating environment using a stream of monocular images. For instance, the harvesting machine identifies and localizes a shake point of a tree by projecting virtual rays from the pose of the identification system to the identified emergence point feature. To harvest the fruit of trees in the orchard, the harvesting machine shakes the tree at the identified shake point. Additionally, the harvesting machine autonomously navigates through the orchard using a combination high resolution spatial information based on localized features and low resolution spatial information from accessed satellite images.