AQ-PCDSys: An Adaptive Quantized Planetary Crater Detection System for Autonomous Space Exploration

TL;DR

AQ-PCDSys tackles the challenge of real-time crater detection under the strict memory and power constraints of space-qualified onboard hardware. It integrates a Quantization Aware Backbone with an Adaptive Multi-Sensor Fusion module and Multi-Scale Detection Heads to fuse Optical Imagery and DEM data, maintaining high fidelity while operating on INT8 arithmetic. The approach includes a specialized loss framework with a small-crater loss boost and a hardware-aware training-deployment pipeline, aiming to deliver robust, low-latency crater detection for autonomous landing, navigation, and hazard avoidance. The work demonstrates a principled co-design that balances accuracy, speed, and power—crucial for future deep-space missions—and outlines concrete steps for empirical hardware validation and broader terrain understanding.

Abstract

Successful autonomous planetary exploration hinges on real-time, high-fidelity environmental perception. However, standard deep learning models usually demand far more memory and computation power than space-qualified, radiation-hardened onboard hardware can provide. This creates a fundamental design challenge of deploying sophisticated detection architectures without saturating the rigid power and memory envelopes of the computation hardware of planetary exploration platforms. We propose the Adaptive Quantized Planetary Crater Detection System to resolve this bottleneck. Our framework integrates a Quantized Neural Network, refined through Quantization Aware Training, with an Adaptive Multi-Sensor Fusion module. By forcing weights into low-precision integer arithmetic, we effectively strip away the floating-point overhead that typically bottlenecks onboard processors and system memory. This yields a leaner model footprint and significantly faster processing while the detection fidelity remains high. Such efficiency enables AMF module to merge high-bandwidth Optical Imagery streams with Digital Elevation Models using an Adaptive Weighting Mechanism to re-balance sensor priority under variable conditions like deep shadows or high albedo. Integrated Multi-Scale Detection Heads then resolve craters across a wide range of diameters, providing a computationally efficient and precise solution for real-time detection, localization of craters and hazard avoidance. This paper establishes the architectural design and theoretical justification of the system. While our methodology is grounded in principles of hybrid computer vision and planetary science, we present this as a blueprint for future empirical validation and hardware benchmarking on integer-arithmetic units. This system provides a capability vital for the next generation of autonomous landing, navigation, and deep space explorations.

Figures (6)

• Figure 1: High-level architecture of the autonomous navigation platform, showing data flow from sensors through the AQ-PCDSys pipeline to mission control systems.
• Figure 2: Block diagram of the AQ-PCDSys Quantization Aware Backbones, Adaptive Multi-Sensor Fusion module, and Multi-Scale Detection Heads.
• Figure 3: Block diagram of Quantization Aware Backbone.
• Figure 4: Detailed block diagram of the Adaptive Multi-Sensor Fusion (AMF) module, illustrating the attention sub-networks for OI and DEM data.
• Figure 5: Flowchart of Adaptive Weighting Mechanism (AWM).