<?xml version="1.0" encoding="utf-8"?>
<export-example>
  <doc>
    <id>872</id>
    <completedYear>2021</completedYear>
    <publishedYear>2020</publishedYear>
    <thesisYearAccepted/>
    <language>deu</language>
    <pageFirst/>
    <pageLast/>
    <pageNumber>41, XVII</pageNumber>
    <edition/>
    <issue/>
    <volume/>
    <articleNumber/>
    <type>masterthesis</type>
    <publisherName/>
    <publisherPlace>Ingolstadt</publisherPlace>
    <creatingCorporation/>
    <contributingCorporation/>
    <belongsToBibliography>0</belongsToBibliography>
    <completedDate>2021-03-22</completedDate>
    <publishedDate>2020-09-16</publishedDate>
    <thesisDateAccepted>--</thesisDateAccepted>
    <title language="deu">Validation of a CNN classifier for RADAR in real and simulation domain, exclusively trained in simulated RSI model</title>
    <abstract language="deu">The current developments in driver assistance systems enabled OEMs to oﬀer vehicles, ready for highly automatic driving (Level 3). A key factor for this development is the environmental perception, based on sensor-systems. Currently, LIght Detection And Ranging (LIDAR), Cameras and RAnge Detection And Ranging (RADAR) are the mainly used sensor-systems, to generate a perception of the vehicular environment. Due to diﬀerent used spectrums, the performance and reliability of a sensor depends on environmental conditions, like darkness, rain or fog. RADAR is known as a robust sensor, working well also in rain or light conditions. A key feature for an object, surrounding the own car, is a classiﬁcation, independently of a certain sensor-system. It enables the adjustment of the own trajectory to speciﬁc situations, deﬁned by the surrounded objects. Nevertheless, it is challenging to classify an object with RADAR. Common sophisticated automotive classiﬁcation algorithms rely on camera. But a precise classiﬁcation of objects, detected by RADAR, gained on importance due to the reliable characteristics of RADAR mentioned before.&#13;
&#13;
In addition to real world domain, simulation software provides tools for RADAR measurements and deployment. Even physical sensor models, near to reality, are implemented in software, enabling the improvement of signal processing techniques in a reproducible, convenient and fast manner. RADAR-speciﬁc eﬀects, like multipath or clutter, are also implemented. The eﬀects cause so-called "Ghost-Targets", pretending fake objects.&#13;
&#13;
This thesis will make a contribution to the question how the simulation of RADAR can be used to classify objects in real world. The classiﬁcation task is taken over by a CNN classiﬁer. Due to the fact that it is a supervised learning method, the classiﬁer gets trained. Current published papers used real but quit static data for the training. This thesis uses data out of IPG Carmaker 8 simulation model.&#13;
&#13;
The proposed methods cluster the data with a state of the art cluster algorithm for RADAR and label, based on ground truth, with a novel approach. A multi-class CNN, implemented in Tensorflow, is enabled to diﬀer between a car, bicycle or pedestrian. The evaluation is concentrated on a confusion matrix. Beside simulation data, also real world data is used to evaluate the classiﬁer. Therefore, a scenario in real world is created. The measurement is performed with INRAS RadarLog, specialized for research and raw data analyze. Both domains, simulation and real world, are scaled on a common frame size and amplitude. The comparison should indicate the potential of a classiﬁer to classify objects with a certain accuracy. The appropriate classiﬁer is exclusively trained in simulation domain and gets applied in the real world and simulation domain. Weather phenomenons like rain, fog or extreme temperatures are excluded in the simulation.</abstract>
    <identifier type="urn">urn:nbn:de:bvb:573-8727</identifier>
    <enrichment key="opus.import.date">2021-03-22T10:46:02+00:00</enrichment>
    <enrichment key="opus.source">sword</enrichment>
    <enrichment key="opus.import.user">primuss</enrichment>
    <licence>Urheberrechtsschutz</licence>
    <advisor>
      <first_name>Werner</first_name>
      <last_name>Huber</last_name>
    </advisor>
    <author>
      <first_name>Marinus</first_name>
      <last_name>Wollschlager</last_name>
    </author>
    <advisor>
      <first_name>Michael</first_name>
      <last_name>Botsch</last_name>
    </advisor>
    <collection role="Import" number="import">Import</collection>
    <collection role="institutes" number="19311">Fakultät Elektro- und Informationstechnik</collection>
    <collection role="degree_programme" number="19348">Automatisiertes Fahren und Fahrzeugsicherheit (M. Eng.)</collection>
    <thesisPublisher>Technische Hochschule Ingolstadt</thesisPublisher>
    <thesisGrantor>Technische Hochschule Ingolstadt</thesisGrantor>
    <file>https://opus4.kobv.de/opus4-haw/files/872/I000769947Abschlussarbeit.pdf</file>
  </doc>
</export-example>
