Aiming at the problems of large amount of data collection, low collection efficiency and binding experimental objects existing in the current insect flight information collection methods, an automatic collection device based on visual inspection technology was designed. The device mainly included obstacle channel module, motion trigger acquisition module, flight path information module and automatic acquisition control system. Firstly, gap controller was designed according to the requirements of the control function, and closed-loop control was realized by using hardware such as laser ranging sensor and stepping motor; secondly, the STM32 microcontroller was used as control terminal, and the Blackfly S USB3 high-speed camera was triggered by using motion direction detection algorithm to realize the sequence image acquisition of insects in the specified motion direction; then, using the open-source computer vision library OpenCV to analyze the collected sequence images and obtain the flight path information of insects; finally, the embedded control system coordinated the communication between modules to achieve the design requirements of efficient and stable acquisition of insect flight information. In order to verify the applicability and accuracy of the device, Apis cerana was selected for the gap crossing experiment. The experimental results showed that in a static environment, the device could obtain complete and clear flight sequence images. The average accuracy rate reached 73.24%. The acquisition performance was stable, which effectively improved the acquisition efficiency. By analyzing the flight path information of honeybee, it is inferred that the mechanism of gap recognition is related to the amplitude and speed of its lateral movement, which provides support for in-depth research on the flight mechanism of Apis cerana.