This thesis examines the context effects on retrieval, and the influence of action on the representation of distance in cognitive maps. It is proposed that bias in distance estimation is a function of the contexts of retrieval that trigger the representation of action in memory during evaluation tasks. The proposal is consistent with embodied cognition evidence that suggests that actions are implicitly a part of the representation, and will be naturally extracted as part of the retrieval process. The experimental work presented examines two different contextual cues; the frequency of visitation to landmarks, and the importance of activity performed at landmarks. Each cue primes differently the conceptualisation of landmarks prior to making distance estimation. This priming facilitates memory access, which fleshes out relevant spatial information from cognitive maps that are used in distance estimation and route description. This proposal was examined in a series of four experiments that employed structured interviews. Participants had to rate landmarks based on frequency of visitation criteria or importance of activity criteria, or both. They then made verbal distance estimations and route descriptions. The results found implicate the involvement of action representation. The involvement of action in cognitive process was empirically investigated in three further experiments. A new methodology was developed featuring the use of a blindfold, linguistic descriptions, and control of actual movements. Blindfolded participants learned new environments through verbal descriptions by imagining themselves walking in time with the metronome beats. During turns, they were carefully moved. Following instructions, they performed an action at mid-route. Their memories for the newly learned environments were tested through recalls and measured again with the metronome beats. The results found were consistent with explanations based on network-map theory. They implicate attentional processes as an intrinsic part of the cognitive mechanism, and the strings of the network-map as the actual motor program that executes the movement. These results are discussed in relation to the nature of cognitive maps.

This master thesis investigates the problem of making an unmanned aerial vehicle(uav) follow a person or a group of persons while keeping a fixed distance tothe chosen target. The purpose of this thesis is to give a proof of concept prototypeof how such a system would work to achieve that task. The main problemconsists of controlling the uav based on visual input from a camera. With thehelp of a visual object detection and tracking system, the image coordinates ofthe targets can be found. An algorithm was developed to calculate the target’sworld position based on its image coordinates and the world position and orientationof the uav. A control system was implemented that uses that uses thepositional information to set the velocity of the uav, if its position needs to bechanged. Several strategies for handling groups of targets were investigated. Inaddition a simulator was developed that can be used to simulate the image coordinatesof a target when the world position of the target is known. The systemwas tested during live flights, using a high precision motion capture system forreference. The results were mainly positive in showing proof of concept and evenshowing a relatively high level of precision.

Our research focusses on the problem of finding nearby peers in the Internet. We focus on one particular approach, Beacon Triangulation that is widely used to solve the peer-finding problem. Beacon Triangulation is based on relative distances of nodes to some special nodes called beacons. The scheme gives an error when a new node that wishes to join the network has the same relative distance to two or more nodes. One of the reasons for the error is that two or more nodes have the same distance vectors. As a part of our research work, we derive the conditions to ensure the uniqueness of distance vectors in any network given the shortest path distribution of nodes in that network. We verify our analytical results for G(n, p) graphs and the Internet. We also derive other conditions under which the error in the Beacon Triangulation scheme reduces to zero. We compare the Beacon Triangulation scheme to another well-known distance estimation scheme known as Global Network Positioning (GNP).

Thesis (Ph. D.)--Miami University, Dept. of Psychology, 2006. / Title from second page of PDF document. Includes bibliographical references (p. 64-72).

No description available.

The main purpose of this thesis is to utilize landmarks with known coordinates and the distance between a target and landmarks to establish an objective function, and to optimize the objective function by using unconstrained direct search method to estimate the coordinate of target. A number of nodes in the sensor network serve as the landmarks according to landmark selection algorithm. Since the landmark selection algorithm is time-consuming, a simplified scheme that would improve the algorithm is to reuse the distance information that had been computed. Due to the limit of transmission range between nodes, utilizing the shortest-path distance estimation model can quickly estimate the distance between the target and non-adjacent landmarks. The main conception of the model is combining the manner of multi-hop with the shortest-path model. Due to the possible errors in distance estimation, the error per hop is considered for reducing the estimation errors. It will obviously reduce the localization errors of the target. The thesis utilizes unconstrained direct search method to optimize the objective functions such as the simplex evolutionary method (SEM), the cyclic coordinate method(CCM) and the Powell method (PM). CCM and PM will tackle the problem of finding the forward length along search direction. Hence, two schemes that combine CCM or PM with SEM are proposed to resolve the problem. Finally, simulations are conducted to generate random some nodes in an known area and to select landmarks from the nodes. Let the target be assigned in the area and do performance analysis of positioning algorithm. We discuss the performance of the positioning algorithm by considering the error per hop approach. We also discuss the effects on positioning by changing some variables such as the number of nodes, the number of landmarks and the transmission range of nodes. It is seen that the positioning errors will be reduced in examples where the number of landmark are four or the number of node are four hundred. The performance of positioning becomes accurate by reducing the distance estimation error.

The object with this project is to develop a more accurate way to measure the level in the fuel tank in Scaniavehicles. The level should be displayed for the driver and a warning system be implemented to make the driveraware if the fuel level is too low. Furthermore a main goal is to develop an estimation of the distance that thevehicle could travel before refueling is needed.The fuel level estimation system is modeled using Matlab Simulink and simulated with measurement datacollected from real driving scenarios. After evaluating the system it is implemented in one of the electricalcontrol units located on a test vehicle which communicates with other systems. After implementation more testsare performed with the test vehicle to verify that the same functionality achieved during simulations is achievedusing the system implemented in a vehicle.The fuel level estimated with a KF (Kalman filter) that uses fuel consumption and level measurement results ingood performance. A more stable level estimate is achieved and a negative elevation of the estimate most of thetime, as a result of fuel use. Compared to the method Scania vehicles estimate their fuel level with today thenew level estimate is more steady and not that easily affected by fuel movements. The KF is more demanding interms of memory allocation, processor speed and inputs needed, which has to be considered when comparingboth methods. Another disadvantage with the KF is that it is dependent on the samples from the fuel levelsensor to get an initial estimate during startup.Furthermore the KF is easily expanded with more inputs that use information from other sensors on other parts of the vehicle.

<p>The object with this project is to develop a more accurate way to measure the level in the fuel tank in Scaniavehicles. The level should be displayed for the driver and a warning system be implemented to make the driveraware if the fuel level is too low. Furthermore a main goal is to develop an estimation of the distance that thevehicle could travel before refueling is needed.The fuel level estimation system is modeled using Matlab Simulink and simulated with measurement datacollected from real driving scenarios. After evaluating the system it is implemented in one of the electricalcontrol units located on a test vehicle which communicates with other systems. After implementation more testsare performed with the test vehicle to verify that the same functionality achieved during simulations is achievedusing the system implemented in a vehicle.The fuel level estimated with a KF (Kalman filter) that uses fuel consumption and level measurement results ingood performance. A more stable level estimate is achieved and a negative elevation of the estimate most of thetime, as a result of fuel use. Compared to the method Scania vehicles estimate their fuel level with today thenew level estimate is more steady and not that easily affected by fuel movements. The KF is more demanding interms of memory allocation, processor speed and inputs needed, which has to be considered when comparingboth methods. Another disadvantage with the KF is that it is dependent on the samples from the fuel levelsensor to get an initial estimate during startup.Furthermore the KF is easily expanded with more inputs that use information from other sensors on other parts of the vehicle.</p>

Currently, as one of the most popular technologies being discussed and experimented, the application of flying copters in different industries is facing an obvious barrier; which is how to avoid obstacles while flying. One of the industries among all is small-sized package delivery business, which is also the master topic of a series of experiments. The most popular designs that have used for the Flying Copter Obstacle Avoidance System such as lidar scanners and infrared rangefinders are significantly accurate. However, with the heavyweight, expensive price and higher power consumption, these systems cannot be put into mass production. To reduce the cost and power consumption of the Obstacle Avoidance System, an innovative vision-based low-cost Obstacle Distance Estimation System for flying copters is demonstrated in this thesis. The Fisheye Lens Camera is used to provide a broader detection range and accurate results. Compared to other standard vision-based systems, the Fish Lens Camera Distance Estimation System can provide (around 360 degrees) extensive view for obstacle detection. Through the parallax pictures captured by the camera and the trigonometric rules, the system can estimate the distance to the target obstacle with reasonable results.

<p>  </p> <p>The maintenance of a healthy sewer infrastructure is a major challenge due to the root damages from nearby plants that grow through pipe cracks or loose joints, which may lead to serious pipe blockages and collapse. Traditional inspections based on video surveillance to identify and localize root damages within such complex sewer networks are inefficient, laborious, and error-prone. Therefore, this study aims to develop a robust and efficient approach to automatically detect root damages and localize their circumferential and longitudinal positions in CCTV inspection videos by applying deep neural networks and computer vision techniques. With twenty inspection videos collected from various resources, keyframes were extracted from each video according to the difference in a LUV color space with certain selections of local maxima. To recognize distance information from video subtitles, OCR models such as Tesseract and CRNN-CTC were implemented and led to a 90% of recognition accuracy. In addition, a pre-trained segmentation model was applied to detect root damages, but it also found many false positive predictions. By applying a well-tuned YoloV3 model on the detection of pipe joints leveraging the Convex Hull Overlap (<em>CHO</em>) feature, we were able to achieve a 20% improvement on the reliability and accuracy of damage identifications. Moreover, an end-to-end deep learning pipeline that involved Triangle Similarity Theorem (<em>TST</em>) was successfully designed to predict the longitudinal position of each identified root damage. The prediction error was less than 1.0 feet. </p>