1,720,965 research outputs found
Gamma-Induced Image Degradation Analysis of Robot Vision Sensor for Autonomous Inspection of Nuclear Sites
Full text linkThere is an increasing desire to conduct autonomous inspection of nuclear sites using robots. However, the presence of gamma radiation in nuclear sites induces degradation in vision sensors. In this paper, the effects of gamma radiation on a robot vision sensor (CMOS camera) used for radiological inspection is examined. The analyses have been carried out for two types of images at different dose rates: a) dark images b) illuminated images. In this work, dark images and chessboard images under illumination are analysed using various evaluation metrics to evaluate the effect of gamma radiation on CMOS Integrated Circuit (IC) and electronic circuitry of the sensor. Experimental results manifest significant changes in electrical properties like the generation of radiation-induced photo signal in sensing circuitry and radiation-induced noise affecting the visual odometry of the robot. System-level degradation for gamma dose rates upto 3 Gy/min intensifies, making data from the imaging sensor unreliable for the visual odometry. However, images captured for gamma dose rate upto 3 Gy/min can be used for surveillance purpose
Coverage Path Planning Techniques for Inspection of Disjoint Regions with Precedence Provision
Full text linkRecent times are witnessing an emergence of sites that are hazardous for human access. This has created a global demand to equip agents with the ability to autonomously inspect such environments by computing a coverage path effectively and efficiently. However, inspection of such sites requires agents to consider the correlation of work, providing precedence provision in visiting regions. The current approaches to compute coverage path in the hazardous sites, however, do not consider precedence provision. To this end, coverage path planning strategies are proposed, which provide precedence provision. To meet the challenges, the problem is divided into two phases: inter-region and intra-region path planning. In the ‘inter-region’ path planning of the approach, the site comprising of multiple disjoint regions is modelled as connectivity graph. Two novel approaches, Mixed Integer Linear Programming (MILP) solution and heuristic based techniques, are proposed to generate the ordered sequence of regions to be traversed. In the ‘intra-region’ path planning of the approach, each region is decomposed into a grid and Boustrophedon Motion is planned over each region. The ability of combined approach to provide complete coverage is proved under minor assumption. An investigative study has been conducted to elucidate the efficiency of the proposed approach in different scenarios using simulation experiments. The proposed approach is evaluated against baseline approaches. The results manifest a significant reduction in cost and execution time, which caters to inspection of target sites comprising of multiple disjoint regions with precedence provision
Weather Classification: A new multi-class dataset, data augmentation approach and comprehensive evaluations of Convolutional Neural Networks
Full text linkWeather conditions often disrupt the proper functioning of transportation systems. Present systems either deploy an array of sensors or use an in-vehicle camera to predict weather conditions. These solutions have resulted in incremental cost and limited scope. To ensure smooth operation of all transportation services in all-weather conditions, a reliable detection system is necessary to classify weather in wild. The challenges involved in solving this problem is that weather conditions are diverse in nature and there is an absence of discriminate features among various weather conditions. The existing works to solve this problem have been scene specific and have targeted classification of two categories of weather. In this paper, we have created a new open source dataset consisting of images depicting three classes of weather i.e rain, snow and fog called RFS Dataset. A novel algorithm has also been proposed which has used super pixel delimiting masks as a form of data augmentation, leading to reasonable results with respect to ten Convolutional Neural Network architectures
Weather Classification by Utilizing Synthetic Data.
Full text linkWeather prediction from real-world images can be termed a complex task when targeting classification using neural networks. Moreover, the number of images throughout the available datasets can contain a huge amount of variance when comparing locations with the weather those images are representing. In this article, the capabilities of a custom built driver simulator are explored specifically to simulate a wide range of weather conditions. Moreover, the performance of a new synthetic dataset generated by the above simulator is also assessed. The results indicate that the use of synthetic datasets in conjunction with real-world datasets can increase the training efficiency of the CNNs by as much as 74%. The article paves a way forward to tackle the persistent problem of bias in vision-based datasets
An Offline-Online Strategy for Goal-Oriented Coverage Path Planning using A Priori Information
No full textRecent times are witnessing the emergence of indoor sites with extenuating circumstances that place a strict time constraint on mobile robots to reach a target while covering a given area. This has created a global demand to equip mobile robots with the ability to autonomously plan a coverage path to reach the static target effectively and efficiently. The current approaches to achieve such tasks, however, are either time-consuming or human-operator dependent. To this end, an offline-online strategy is proposed to meet the speeding-up challenge by efficiently modelling the environment using a priori information. In the ‘offline’ stage of the strategy, the layout of the environment is segmented into a set of regions. The corners and dead-ends are identified based on the spatial mobility of the regions. The global path is then computed by deriving a graph-structured, road map using the segmented regions. In the ‘online’ stage, the global path is traversed by selecting frontiers which concurrently minimizes the covered area and time. In case the path is obstructed, a re-planning strategy is deployed. The proposed strategy is evaluated by various experiments against two baseline search approaches in three simulated environments. The results manifest a significant reduction in time to reach the goal and coverage area which caters to the strict time constraint for mobile robot
Gamma-induced Degradation Analysis of Commercial off-the-shelf Camera Sensors
Full text linkIn this work, we investigated the degradation of commercial camera sensors exposed to 100 Gy of γ dose at dose rates of 0.55 Gy/min and 1.34 Gy/min respectively. The results show that the degradation is strongly dependent on the dose rate but doesnt vary much with the accumulation of dose at constant dose rate. Furthermore, cameras with in-built processing electronics are more susceptible to gamma radiations as compared to the cameras with sensing unit only
Degradation Measurement of Kinect Sensor Under Fast Neutron Beamline
Full text linkThe neutron-induced degradation of depth images acquired using a Kinect sensor is investigated through the standard metrics. The evaluated metrics indicate saturation of degradation after several hours of exposure, which is recovered through annealing
Coverage Path Planning for Autonomous Robots
Full text linkCoverage Path Planning (CPP) is a problem of path computation with minimal length that guarantees to scan the entire area of interest. CPP finds its application in diverse fields like cartography, inspection, precision agriculture, milling, and demining. However, this thesis is a prominent step to solve CPP for real-world problems where environment poses multiple challenges. At first, four significant and pressing challenges for CPP in extreme environment are identified. Each challenge is formulated as a problem and its solution has been presented as a dedicated chapter in this thesis. The first problem, Goal-Oriented Sensor based CPP, focuses on cumbersome tasks like Nuclear Decommissioning, where the robot covers an abandoned site in tandem with the goal to reach a static target in minimal time. To meet the grave speeding-up challenge, a novel offline-online strategy is proposed that efficiently models the site using floor plans and grid maps as a priori information. The proposed strategy outperforms the two baseline approaches with reduction in coverage time by 45%- 82%. The second problem explores CPP of distributed regions, applicable in post-disaster scenarios like Fukushima Daiichi.
Experiments are conducted at radiation laboratory to identify the constraints robot would be subjected to. The thesis is successfully able to diagnose transient damage in the robot’s sensor after 3 Gy of gamma radiation exposure. Therefore, a region order travel constraint known as Precedence Provision is imposed for successful coverage. The region order constraint allows the coverage length to be minimised by 65% in comparison to state-of-the-art techniques. The third problem identifies the major bottleneck of limited on-board energy that inhibits complete coverage of distributed regions. The existing approaches allow robots to undertake multiple tours for complete coverage which is impractical in many scenarios. To this end, a novel algorithm is proposed that solves a variant of CPP where the robot aims to achieve near-optimal area coverage due to path length limitation caused by the energy constraint. The proposed algorithm covers 23% - 35% more area in comparison to the state-of-the-art approaches. Finally, the last problem, an extension of the second and third problems, deals with the problem of CPP over a set of disjoint regions using a fleet of heterogeneous aerial robots. A heuristic is proposed to deliver solutions within acceptable time limits. The experiments demonstrate that the proposed heuristic solution reduces the energy cost by 15-40% in comparison to the state-of-the art solutions
- …
