In this paper, we design a technologically intelligent wheelchair with eye-movement control for patients with ALS in a natural environment. The system consists of an electric wheelchair, a vision system, a two-dimensional robotic arm, and a main control system. The smart wheelchair obtains the eye image of the controller through a monocular camera and uses deep learning and an attention mechanism to calculate the eye-movement direction. In addition, starting from the relationship between the trajectory of the joystick and the wheelchair speed, we establish a motion acceleration model of the smart wheelchair, which reduces the sudden acceleration of the smart wheelchair during rapid motion and improves the smoothness of the motion of the smart wheelchair. The lightweight eye-movement recognition model is transplanted into an embedded AI controller. The test results show that the accuracy of eye-movement direction recognition is 98.49%, the wheelchair movement speed is up to 1 m/s, and the movement trajectory is smooth, without sudden changes.

Blind and visually impaired (BVI) people face significant challenges in perception, navigation, and safety during travel. Existing infrastructure (e.g., blind lanes) and traditional aids (e.g., walking sticks, basic audio feedback) provide limited flexibility and interactivity for complex environments. To solve this problem, we propose a real-time travel assistance system based on deep learning. The hardware comprises an NVIDIA Jetson Nano controller, an Intel D435i depth camera for environmental sensing, and SG90 servo motors for feedback. To address embedded device computational constraints, we developed a lightweight object detection and segmentation algorithm. Key innovations include a multi-scale attention feature extraction backbone, a dual-stream fusion module incorporating the Mamba architecture, and adaptive context-aware detection/segmentation heads. This design ensures high computational efficiency and real-time performance. The system workflow is as follows: (1) the D435i captures real-time environmental data; (2) the processor analyzes this data, converting obstacle distances and path deviations into electrical signals; (3) servo motors deliver vibratory feedback for guidance and alerts. Preliminary tests confirm that the system can effectively detect obstacles and correct path deviations in real time, suggesting its potential to assist BVI users. However, as this is a work in progress, comprehensive field trials with BVI participants are required to fully validate its efficacy.

Blind and visually impaired (BVI) people face multiple challenges in terms of perception, navigation and safety when traveling. Although there are infrastructures such as blind alleys and pavement markings in cities, the design and maintenance of these facilities are often inadequate to fully meet the needs of traveling. Traditional travel aids usually rely on crutches or simple acoustic feedback, with poor flexibility and interactivity, making it difficult to effectively cope with complex travel environments. Therefore, it is especially necessary to design a set of intelligent and real-time travel assistive devices. Based on this, this paper proposes a set of travel assistance devices based on deep learning. The hardware part of the device includes the main controller-NVIDIA Jetson Nano, the environment sensing unit-D435i depth camera and the feedback unit-SG90 servo. In the software part, in order to meet the limitation of the embedded device's arithmetic power, this paper designs a lightweight target detection and segmentation algorithm to realize obstacle detection and forward direction guidance, which mainly includes a multi-scale attention feature extraction backbone network, a dual-stream fusion module combined with the Mamba architecture, and detection and segmentation heads capable of adaptive context-awareness. The algorithm has high topicality and computational efficiency, which makes it possible to meet the needs of blind people traveling on low-power embedded devices. The overall workflow of the system is as follows: firstly, the binocular depth camera D435i captures real-time information about the surrounding environment; then, the processor recognizes and analyzes the acquired data and converts the obstacle distance and road direction offset signals into Arduino electrical signals, and finally, the servo provides the vibration feedback in order to guide and warn the blind. After the field test, the results show that the device can help the blind find and avoid obstacles in time, correct the traveling position deviation, and meet the real-time requirements of the blind traveling.

It is particularly important to study and analyze the airport passenger volume and its influencing factors for the development of China's aviation industry and tourism industry. Based on BP neural network, the passenger volume prediction model of Xinzheng International Airport is established, and through the analysis and summary of the existing literature, many factors affecting the airport freight volume are analyzed, and four main influencing factors are selected as the input layer, and then the model is trained by matlab software. It can be concluded from the results that the maximum prediction error rate is less than 5%, which shows that the model used in this paper has high accuracy, and can provide a new way for airport freight volume prediction, so as to improve the forecasting accuracy of airport freight volume in China and provide reference for local airport construction and development planning.