As a large-scale precision electromechanical product, the assembly process of civil airplanes requires extremely strict monitoring of excess material. At present, in the field of aircraft assembly, the detection of excess material still relies on manual visualization, which is less efficient and cumbersome, and has certain limitations in terms of accuracy and efficiency. Especially in the hidden dark area of aircraft assembly, the traditional method is often difficult to achieve the expected detection effect. Therefore, the development of a hidden environment redundancy detection system based on intelligent terminals is of great significance to improve aircraft flight safety and assembly efficiency. In this paper, a set of aircraft assembly hidden environment FOD (Foreign Object Debris) intelligent detection system is constructed based on machine learning and visual detection technology with industrial endoscope and intelligent terminal as the hardware carrier; according to the compilation environment of Android Studio, a software for detecting redundant objects of civil aircraft is designed and developed, which realizes accurate positioning and type identification of redundant objects under the hidden environment. The system realizes the precise positioning and type recognition of excess objects in the hidden environment.