O'Keefe, Kyle P. G.Naghdi, Sharareh2020-07-102020-07-102020-07-09Naghdi, S. (2020). Improving Bluetooth-based Indoor Positioning Using Vision and Artificial Networks (Doctoral thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca.http://hdl.handle.net/1880/112282The demands for accurate positioning and navigation applications in complex indoor environments such as emergency call positioning, fire-fighting services, and rescue operations are increasing continuously. Global Navigation Satellite Systems (GNSS) receivers, while ubiquitous in outdoor positioning, are not effective indoors. One of the best solutions to solve this problem and increase the accuracy of the user's position in indoor areas is to apply other sensors. This research takes advantage of Bluetooth Low Energy (BLE) technology, vision systems, and Artificial Neural Networks (ANNs) to improve the accuracy of the position solutions in indoor environments for pedestrian applications. BLE technology faces challenges related to the Received Signal Strength Indicator (RSSI) fluctuations caused by human body shadowing. This thesis presents methods to compensate for losses in the RSSI values by applying ANN algorithms to RSSI measurements from three BLE advertising channels. The resulting improved RSSI values are then converted into ranges using path loss models and trilateration is applied to obtain indoor positions. Two neural network algorithms were implemented. The first used only the RSSI values while the second incorporated a wearable camera as an additional source of information about the presence or absence of human obstacles. The results showed that the two proposed artificial-based systems could enhance RSSI due to human body shadowing and provide significantly better ranging and positioning solutions than fingerprinting and trilateration techniques with uncorrected RSSI values. Two proposed systems provided 3.7 m and 6.7 m positioning accuracy in 90 % of the time in a complex environment with the presence of the human body, nevertheless, the fingerprinting and the classic algorithms offered 8.7 m and 12.3 m position accuracy in the same situation, respectively.University of Calgary graduate students retain copyright ownership and moral rights for their thesis. You may use this material in any way that is permitted by the Copyright Act or through licensing that has been assigned to the document. For uses that are not allowable under copyright legislation or licensing, you are required to seek permission.Engineering-GeomaticsIndoor positioningBluetooth Low Energy (BLE)Neural NetworksTrilateration techniquesSeparate advertising channelsWearable cameraHuman body detectionReceived signal strengthShadowing effectEngineeringEngineering--Electronics and ElectricalRoboticsdoctoral thesis10.11575/PRISM/38004