Design And Implementation Of Indoor Positioning System Based On Ceiling

With the development of society,inner location services based on moving position which plays an important role in various industries and people's daily lives.The rise of projects such as smart homes,robots,and autonomous driving has driven the demand for precise indoor positioning to become stronger.In the technical field,indoor positioning technologies are mainly based on radio frequency,Bluetooth,and sound waves,but vision-based indoor positioning technologies are not as mature as other technologies.Visual positioning has the advantages of easy to get equipment,low cost,and full use of image information,so it has broad development prospects.Vision-based indoor positioning starts with effective pixels which obtain from the image.In this paper,with the help of natural beacons such as ceilings,the coordinates of the four top corners of a ceiling are used as the coordinates of the image points through Hough line segment detection.To obtain the object point coordinates corresponding to the image point,way can be done to bring the image point into the homography matrix.Finally,the coordinates of the image point and the object point are brought into the calculation program of the rear intersection,and the spatial coordinates of the corresponding lens can be obtained.In order to achieve the above solutions,some challenges encountered in this paper :(1)The captured ceiling image contains a lot of noise,which leads to errors and noisy straight lines in the line detection effect;(2)A false peak appeared near the real peak in Hough space,which resulted in parallel and close line segments in the result,and also caused a disconnection in the middle of a long line segment;(3)How to roughly locate the position of the ceiling in the image through image processing or other methods,and how to identify the corner of the wall,etc.The work was done as follows in this article:(1)Compared with the method of establishing object feature database for indoor positioning,this article innovatively proposes a homography matrix + rear intersection model for indoor positioning,and successfully validates the model theoretically,proving the feasibility of the model;(2)Careful investigation and scientific analysis of the Probabilistic Hough Transform line segment detection method.To improve the Probabilistic Hough Transform line segment detection algorithm,image block and method of line segment fusion method were used.What's more,the new line segment detection algorithm was used as the center of the ceiling line segment detection in this project algorithm;(3)Decoupling the functions of different modules,and finally using a programming language to implement the indoor positioning system.After the above work,the experimental data in this paper show that the positioning accuracy of this scheme is 20 cm,which can be used as an innovation of the monocular vision positioning scheme and provide ideas for further research on indoor positioning.