The Methodology Of Monocular Distance Measurement Based On Eye Location And ARM Platform Implementation

The monocular vision measurement is a non-contact,fast and efficient measurement and detection technology.With the development of computer vision,image processing and pattern recognition technology,the technology of Visual measurement is widely used in the field of robots and security.For the problem of robot collaborative work with human and the demand of measurement positioning to target person during intelligent security monitoring process.This paper put forward a method of monocular distance measurement based on eye location to detect the distance between the target person and camera lens.To begin with,this method collects face images through the USB camera and detects the pupil in the collected face images,then uses Small hole imaging triangular similarity principle and Combines it and the actual pupil distance and camera calibration parameters to calculate the distance between the target eye and camera.At last,this method is implemented based on the embedded ARM platform.The main works of this paper are as follows:1)This paper proposes a method of human eye positioning based on feature and appearance,and this method combines the feature-based and apparent-based methods to detect the human eye,it improves the accuracy of human eye detection and detection of robustness.Firstly,an improved hybrid projection function is proposed to segment the human eye region on the basis of the classical projection method,And then uses the principal component analysis training human eye classifier to classify the human eye area to obtain the only left and right eye area respectively.Lastly,the improved hybrid projection function and the improved fast radial symmetry transform algorithm are used to locate the pupil finally.2)Introduces the camera linear model and the camera calibration based on 2D plane target,and a method of monocular distance measurement based on eye location is proposed by combining it with simple and efficient features of monocular distance measurement.This method collects face images through the USB camera and detects the pupil in that face images in the first,then use the small hole imaging triangular similarity principle and combines it with the actual pupil distance and camera calibration parameters to calculate the distance between target eye and the camera.3)Designs a monocular distance measurement system based on the embedded ARM platform,and to realize the method of monocular distance measurement based on human eye Positioning.Experiments show that the method proposed in this paper has a high accuracy and robustness.