Advanced machine vision and deep learning models are now widely used as virtual sensors in various applications to detect and classify objects in image measurements. Typically, these virtual sensors output measurement...
详细信息
ISBN:
(纸本)9798350371420;9781737749769
Advanced machine vision and deep learning models are now widely used as virtual sensors in various applications to detect and classify objects in image measurements. Typically, these virtual sensors output measurements as a set of class probabilities for each detected object within the sensor's field of view. However, integrating this type of data into multi-target tracking systems, traditionally based on point measurement detections, presents some challenges. This paper proposes a solution by introducing a labeled multi-Bernoulli filter formulated to handle detections with class measurements from virtual sensors for multi-target tracking. Furthermore, we explore the application of this filter for multi-sensor multi-target tracking within a distributed sensor network. Through numerical experiments involving vehicles equipped with cameras and deep learning classification modules navigating complex roads, we demonstrate that incorporating class information into the tracking process improves tracking accuracy. This improvement is further enhanced when vehicles share and fuse their information in a distributed manner. Our findings highlight the benefits of integrating class probability data into multi-target tracking filters, offering substantial improvements in the accuracy of automated monitoring systems in dynamic and complex environments.
The object-oriented paradigm is characterized in a general sense by a grouping of information with the concept or entity to which it relates. Corresponding to this rather vague definition, there is a wide range of sys...
详细信息
The object-oriented paradigm is characterized in a general sense by a grouping of information with the concept or entity to which it relates. Corresponding to this rather vague definition, there is a wide range of systems that can be classified as object-oriented. However, such systems may provide significantly different perspectives on the structure and manipulation of objects. This stems principally from the different motivations underlying the distinct fields from which object-oriented systems have emerged, such as Data Bases, Artificial Intelligence and Programming Languages. As a result, a myriad of systems have appeared in which diverse terminology is used. For example, are terms such as class, frame, term, actor and entity synonyms, related notions, or descriptions of distinct concepts? How is an object different from these terms? This paper proposes a classification of object-oriented systemsbased upon the conceptualization underlying an object and how such a conceptualization is described. Hence, the issue is more on modelling with objects rather than system idiosyncrasies. Four broad families are identified depending on whether systems follows either a class-based model, a frame-based model, a terminological model or an actor-based model. In the paper, for each category, a definition of the conceptualization is first presented, followed by a description of its characteristics together with some examples of its intended use. The classification serves both as a framework for comparison and as a context within which individual concepts can be described.
暂无评论