An automatic blind guide device based on depth learning algorithm

Abstract

The invention provides an automatic blind guide device based on a depth learning algorithm, which comprises a camera, a navigation device, a speech input module, an ultrasonic module, an infrared module, an IMU inertial measurement module, a depth learning unit and an embedded unit. The output ends of the camera, the navigation device, the voice input module and the IMU inertial measurement moduleare connected with the input ends of the depth learning unit, the output ends of the ultrasonic module and the infrared module are connected with the input ends of the embedded unit, and the output ends of the depth learning unit are connected with the input ends of the embedded unit. The invention realizes the obstacle detection of the surrounding environment of the blind person by means of theultrasonic array worn on the chest of the blind person, and combined with the pressure device driven by the motor, applies a pressure sense corresponding to the obstacle distance to the blind person at the corresponding position of the ultrasonic module, so that the blind people can perceive the distance and orientation of obstacles in the surrounding environment.

Description

technical field [0001] The invention belongs to the technical field, and in particular relates to an automatic blind guide device based on a deep learning algorithm. Background technique [0002] 90% of the information obtained by humans comes from the visual system, therefore, the loss of vision will seriously affect the quality of life. Blind people have great obstacles in their daily life because they cannot see, especially travel difficulties. [0003] There are currently many ways to guide the blind, such as non-visual detection methods such as blind sticks, infrared detection, and ultrasonic detection. However, these methods have certain disadvantages: [0004] ①. Use blind sticks to detect, the detection range is too small, blind people can judge the walking direction by touching the ground or surrounding objects, and whether they need to stop or not; this type of crutches has many disadvantages, especially it is difficult to find distant objects and objects suspen...

AI Technical Summary

Problems solved by technology

[0004] ①. Use blind sticks to detect, the detection range is too small, blind people can judge the walking direction by touching the ground or surrounding objects, and whether they need to stop or not; this type of crutches has many disadvantages, especially it is difficult to find distant objects and objects suspended in mid-air;

[0005] ② Using infrared, ultrasonic and other detection methods can only detect larger obstacles, and cannot obtain complete information such as obstacle contours and heights, so the security is poor

[0006] However, this blind guide reminds the blind through prompt sounds of different volumes, tones, rhythms or segmented voice prompts with large errors, which can only make the blind have a rough judgment on the distance of obstacles, and , consumes a lot of energy of the blind and easily leads to misjudgment

Images