Abstract: The solar system boundary exploration mission faces challenges such as long communication delays, complex environments, and long mission durations, which impose higher requirements on the detector's onboard autonomous data processing and scientific target recognition capabilities. To address these issues, this paper designs and implements an intelligent information processing system. The system adopts the domestic high-reliability aerospace-grade AI chip Yulong810A, combined with a dual-FPGA architecture, and possesses core functions including multi-source information fusion processing, image preprocessing, and target detection. To tackle the challenge of target perception in weak-light complex environments, the system integrates a lightweight target detection algorithm incorporating a low-light enhancement module and introduces a weighted boxes fusion strategy, significantly improving detection accuracy and robustness. The software system employs a layered architecture design, supporting uplink updates, fault-tolerant control, and multi-task cooperative processing. Test results demonstrate that the system performs excellently in terms of target detection accuracy, stability in downlinking engineering parameters, and interrupt handling efficiency, providing key technical support for the solar system boundary exploration mission.