Design And Implementation Of Intelligent Guidance System Based On Brain Computer Interface

In recent years,with the rapid development of embedded technology,wearable portable electronic products have entered the daily life of the public,allowing us to provide certain protection for the daily travel of visually impaired people.Although existing guide devices and products can help,there are some problems that hinder their popularization.In view of this,the project uses brain-computer interface technology to design an intelligent guidance system to bring substantial assistance to visually impaired people.In the research process of the project,the following work was mainly done:(1)The application of the temperature compensation module and brain-computer interface technology,the problem of low reliability of the conventional blinding system has been solved;(2)Using the classification strategy based on Haar-like features and AdaBoost algorithm locates the signal lights in the captured image and uses the OpenCV correlation function to implement the discrimination of the traffic light color.(3)The Canny operator edge detection and Hough transform method are used to extract the parallel lines and achieve the zebra crossing recognition;(4)Using CSP and CELM methods to classify and recognize EEG signals can realize the detection and judgment of the movement of the user's left and right legs and ensure the reliability of the guidance system.Use the Raspberry Pi 3B Development Board as a hardware platform,Python,and OpenCV as software development tools.Image processing and recognition,leg motion judgment,and voice prompts are implemented through the transplantation and configuration of the corresponding interfaces.Through a series of functions and performance tests on the intelligent guiding system,the implementation of various functions of the guiding system is verified.The test results show that increasing the temperature compensation module can ensure the detection accuracy of the obstacle distance;using a large number of samples to train the classifier can improve the classification recognition efficiency,making the system have a certain research significance and practical value.