Development of a Neural Network-based Object Detection for Multirotor Target Tracking PDF Download

Author: Spencer Harwood
Publisher:
ISBN:
Category : Drone aircraft
Languages : en
Pages : 0

View

Book Description

Author: Ashish Kumar
Publisher: Springer Nature
ISBN: 9819932882
Category : Computers
Languages : en
Pages : 280

View

Book Description
With the increase in urban population, it became necessary to keep track of the object of interest. In favor of SDGs for sustainable smart city, with the advancement in technology visual tracking extends to track multi-target present in the scene rather estimating location for single target only. In contrast to single object tracking, multi-target introduces one extra step of detection. Tracking multi-target includes detecting and categorizing the target into multiple classes in the first frame and provides each individual target an ID to keep its track in the subsequent frames of a video stream. One category of multi-target algorithms exploits global information to track the target of the detected target. On the other hand, some algorithms consider present and past information of the target to provide efficient tracking solutions. Apart from these, deep leaning-based algorithms provide reliable and accurate solutions. But, these algorithms are computationally slow when applied in real-time. This book presents and summarizes the various visual tracking algorithms and challenges in the domain. The various feature that can be extracted from the target and target saliency prediction is also covered. It explores a comprehensive analysis of the evolution from traditional methods to deep learning methods, from single object tracking to multi-target tracking. In addition, the application of visual tracking and the future of visual tracking can also be introduced to provide the future aspects in the domain to the reader. This book also discusses the advancement in the area with critical performance analysis of each proposed algorithm. This book will be formulated with intent to uncover the challenges and possibilities of efficient and effective tracking of single or multi-object, addressing the various environmental and hardware challenges. The intended audience includes academicians, engineers, postgraduate students, developers, professionals, military personals, scientists, data analysts, practitioners, and people who are interested in exploring more about tracking.· Another projected audience are the researchers and academicians who identify and develop methodologies, frameworks, tools, and applications through reference citations, literature reviews, quantitative/qualitative results, and discussions.

Author: Marcin Woźniak
Publisher: MDPI
ISBN: 3036512683
Category : Technology & Engineering
Languages : en
Pages : 454

View

Book Description
Recent years have seen a vast development in various methodologies for object detection and feature extraction and recognition, both in theory and in practice. When processing images, videos, or other types of multimedia, one needs efficient solutions to perform fast and reliable processing. Computational intelligence is used for medical screening where the detection of disease symptoms is carried out, in prevention monitoring to detect suspicious behavior, in agriculture systems to help with growing plants and animal breeding, in transportation systems for the control of incoming and outgoing transportation, for unmanned vehicles to detect obstacles and avoid collisions, in optics and materials for the detection of surface damage, etc. In many cases, we use developed techniques which help us to recognize some special features. In the context of this innovative research on computational intelligence, the Special Issue “Advanced Computational Intelligence for Object Detection, Feature Extraction and Recognition in Smart Sensor Environments” present an excellent opportunity for the dissemination of recent results and achievements for further innovations and development. It is my pleasure to present this collection of excellent contributions to the research community. - Prof. Marcin Woźniak, Silesian University of Technology, Poland –

Author: Sheng Jin
Publisher:
ISBN:
Category : Automatic tracking
Languages : en
Pages : 150

View

Book Description

Author: Xianghong (Henry). Liu
Publisher:
ISBN:
Category :
Languages : en
Pages : 154

View

Book Description

Author: Wenbo Wu
Publisher:
ISBN:
Category :
Languages : en
Pages :

View

Book Description
With the rapid development of deep learning techniques, the application of Convolutional Neural Network (CNN) has benefited the task of target object recognition. Several state-of-the-art object detectors have achieved excellent performance on the precision for object recognition. When it comes to applying the detection results for the real world application of collaborative robots, the reliability and robustness of the target object detection stage is essential to support efficient task allocation. In this work, collaborative robots task allocation is based on the assumption that each individual robotic agent possesses specialized capabilities to be matched with detected targets representing tasks to be performed in the surrounding environment which impose specific requirements. The goal is to reach a specialized labor distribution among the individual robots based on best matching their specialized capabilities with the corresponding requirements imposed by the tasks. In order to further improve task recognition with convolutional neural networks in the context of robotic task allocation, this thesis proposes an innovative approach for progressively refining the target detection process by taking advantage of the fact that additional images can be collected by mobile cameras installed on robotic vehicles. The proposed methodology combines a CNN-based object detection module with a refinement module. For the detection module, a two-stage object detector, Mask RCNN, for which some adaptations on region proposal generation are introduced, and a one-stage object detector, YOLO, are experimentally investigated in the context considered. The generated recognition scores serve as input for the refinement module. In the latter, the current detection result is considered as the a priori evidence to enhance the next detection for the same target with the goal to iteratively improve the target recognition scores. Both the Bayesian method and the Dempster-Shafer theory are experimentally investigated to achieve the data fusion process involved in the refinement process. The experimental validation is conducted on indoor search-and-rescue (SAR) scenarios and the results presented in this work demonstrate the feasibility and reliability of the proposed progressive refinement framework, especially when the combination of adapted Mask RCNN and D-S theory data fusion is exploited.

Author: Solomon Negussie Tesema
Publisher:
ISBN:
Category :
Languages : en
Pages : 0

View

Book Description
Object detection is one of the most challenging yet essential computer vision research areas. It means labeling and localizing all known objects of interest on an input image using tightly fit rectangular bounding boxes around the objects. Object detection, having passed through several evolutions and progressions, nowadays relies on the successes of image classification networks based on deep convolutional neural networks. However, as the depth and complication of convolutional neural networks increased, detection speed reduced, and accuracy increased. Unfortunately, most computer vision applications, such as real-time object tracking on an embedded system, requires lightweight, fast and accurate object detection. As a result, object detection acceleration has become a hot research area, with much attention given to FPGA-based acceleration due to FPGA's high-energy efficiency, high-data bandwidth, and flexible programmability.This Ph.D. dissertation proposes incrementally improving object detection models by repurposing existing well-known object detectors into lighter, more accurate, and faster models. Our models achieve a comparable accuracy while being lightweight and faster compared with some of the top state-of-the-art detectors. We also propose and implement object detection inference acceleration using FPGA boards of different capacities and resources. We focus on high resource and energy-efficient inference acceleration implementations while preserving the object detector's accuracy performance. Last but not least, we present various auxiliary contributions such as a highly significant synthetic image generation or augmentation technique for training an object detector which is critical for achieving a high-performance object detector. Overall, our work in this thesis has two parts: designing and implementing lightweight and accurate CPU and GPU-based object detection models and implementing high-throughput, energy, and resource-efficient object detection inference acceleration on an FPGA.

Author: Michael Alan Deadrick
Publisher:
ISBN:
Category : Automatic tracking
Languages : en
Pages : 316

View

Book Description

Author: Dennis Mitzel
Publisher:
ISBN: 9783844025248
Category : Automatic tracking
Languages : en
Pages : 198

View

Book Description
Recent years have seen considerable progress in automotive safety and autonomous navigation applications, fueled by the remarkable advance of individual Computer Vision components, such as object detection, tracking, stereo and visual odometry. The goal in such applications is to automatically infer semantic understanding from the environment, observed from a moving vehicle equipped with a camera system. The pedestrian detection and tracking components constitute an actively researched part in scene understanding, important for safe navigation, path planning, and collision avoidance. Classical tracking-by-detection approaches require a robust object detector that needs to be executed in every frame. However, the detector is typically the most computationally expensive component, especially if more than one object class needs to be detected. A first goal of this thesis was to develop a vision system based on stereo camera input that is able to detect and track multiple pedestrians in real-time. To this end, we propose a hybrid tracking system that combines a computationally cheap low-level tracker with a more complex high-level tracker. The low-level trackers are either based on level-set segmentation or stereo range data together with a point registration algorithm and are employed in order to follow individual pedestrians over time, starting from an initial object detection. In order to cope with drift and to bridge occlusions that cannot be resolved by low-level trackers, the resulting tracklet outputs are fed to a high-level multihypothesis tracker, which performs longer-term data association. With this integration we obtain a real-time tracking framework by reducing object detector applications to fewer frames or even to few small image regions when stereo data is available. Reduction of expensive detector evaluations is especially relevant for the deployment on mobile platforms, where real-time performance is crucial and computational resources are notoriously

Author: Christian Lusardi
Publisher:
ISBN:
Category : Computer vision
Languages : en
Pages : 71

View

Book Description
"Deep Learning allows for great advancements in computer vision research and development. An area that is garnering attention is single object tracking and multi-object tracking. Object tracking continues to progress vastly in terms of detection and building re-identification features, but more effort needs to be dedicated to data association. In this thesis, the goal is to use a graph neural network to combine the information from both the bounding box interaction as well as the appearance feature information in a single association chain. This work is designed to explore the usage of graph neural networks and their message passing abilities during tracking to come up with stronger data associations. This thesis combines all steps from detection through association using state of the art methods along with novel re-identification applications. The metrics used to determine success are Multi-Object Tracking Accuracy (MOTA), Multi-Object Tracking Precision (MOTP), ID Switching (IDs), Mostly Tracked, and Mostly Lost. Within this work, the combination of multiple appearance feature vectors to create a stronger single feature vector is explored to improve accuracy. Different types of data augmentations such as random erase and random noise are explored and their results are examined for effectiveness during tracking. A unique application of triplet loss is also implemented to improve overall network performance as well. Throughout testing, baseline models have been improved upon and each successive improvement is added to the final model output. Each of the improvements results in the sacrifice of some performance metrics but the overall benefits outweigh the costs. The datasets used during this thesis are the UAVDT Benchmark and the MOT Challenge Dataset. These datasets cover aerial-based vehicle tracking and pedestrian tracking. The UAVDT Benchmark and MOT Challenge dataset feature crowded scenery as well as substantial object overlap. This thesis demonstrates the increased matching capabilities of a graph network when paired with a robust and accurate object detector as well as an improved set of appearance feature vectors."--Abstract.