{"id":7483,"date":"2020-02-02T13:54:23","date_gmt":"2020-02-02T12:54:23","guid":{"rendered":"https:\/\/www.l3s.de\/?page_id=7483"},"modified":"2022-12-28T08:00:20","modified_gmt":"2022-12-28T07:00:20","slug":"automatic-damage-detection","status":"publish","type":"page","link":"https:\/\/www.l3s.de\/de\/magazine\/binaire-articles\/automatic-damage-detection\/","title":{"rendered":"Automatische Schadenserkennung"},"content":{"rendered":"
\n \n
\n \n \n
\n \n
\n
\n
\n
<\/div>\n <\/div>\n
\n
<\/div>\n <\/div>\n
\n
<\/div>\n <\/div>\n <\/div>\n
\n \n
\n
\n \n \n
\n \n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n
\n \n
\n \n
\n
\n \n \n
\n \n
\n \n
\n \n
\n
\n \n \n
\n \n
\n \n \n
\n

Ausgabe 01\/2020<\/p>\n <\/div>\n <\/div>\n

\n \n
K\u00fcnstliche Intelligenz<\/span><\/a>\n <\/span>\n <\/div>\n <\/div>\n
\n \n <\/div>\n <\/div>\n <\/div>\n <\/div>\n
\n
\n \n \n
\n \n
\n \n
\n \n \n

Automatische Schadenserkennung<\/h1>\n <\/div>\n <\/div>\n
\n \n <\/div>\n
\n \n
\n \n
\n
\n \n \n
\n \n
\n \n \n
\n

Auf Deutschlands Stra\u00dfen wird es immer enger. In den letzten zehn Jahren stieg die Anzahl der Neuzulassungen f\u00fcr PKW um \u00fcber 30 Prozent. Mit der Zahl der Verkehrsteilnehmer w\u00e4chst auch die Zahl von Schadensf\u00e4llen, die den Versicherungen gemeldet werden. Ausgebildete Mitarbeiter pr\u00fcfen die Daten und Bilder jedes Falles und entscheiden, ob der Einsatz eines Gutachters n\u00f6tig ist. Die Gutachten kosten mehrere tausend Euro und stellen einen nicht unerheblichen Teil des Umsatzes einer Versicherung dar.                                     <\/p>\n

Die automatisierte Bildverarbeitung hat das Potential, diese Aufgabe erheblich zu vereinfachen. In den letzten Jahren hat das Forschungsfeld gro\u00dfe Fortschritte erzielt. Grund daf\u00fcr ist der Einsatz sogenannter Deep-Learning-Systeme, die auf Basis von Daten abstrakte Repr\u00e4sentationen lernen, um verschiedenste Aufgaben zu l\u00f6sen. Am L3S und am Institut f\u00fcr Informationsverarbeitung forscht Prof. Dr. Bodo Rosenhahn mit seinem Team an neuen Verfahren zur automatisierten Bildanalyse. Sie sollen zum Einsatz kommen, um Objekte und deren Eigenschaften in Bildern zu erkennen - so auch Sch\u00e4den an PKW-Bauteilen.<\/p>\n

Gemeinsam mit dem Startup ClaimBuddy<\/a>, einer Ausgr\u00fcndung der Leibniz Universit\u00e4t Hannover auf dem Gebiet des digitalen Schadenmanagements, arbeiten Prof. Rosenhahn und Frederik Schubert in einem F\u00f6rderprojekt daran, die Anforderungen an eine automatisierte Schadenserkennung zu erf\u00fcllen. Dazu z\u00e4hlt der Aufbau eines Datensatzes, um ein Deep-Learning-System zu trainieren, das die Versicherungen bei der Analyse und Bewertung der Schadensf\u00e4lle unterst\u00fctzen kann. F\u00fcr Versicherer bietet sich so die M\u00f6glichkeit einer kostensparenden und zeiteffizienten Regulierung, insbesondere von Kleinsch\u00e4den.<\/p>\n <\/div>\n <\/div>\n

\n \n \n
\n

Abbildung 1<\/p>\n

Beschreibung: Schadenserkennung<\/p>\n

Quelle:\n Car Damage Detective Dataset<\/a>\n <\/p>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n

\n \n \n
\n \n
\n \n
\n \n
\n
\n \n \n
\n \n
\n <\/span><\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n
\n \n
\n \n
\n
\n \n \n
\n \n
\n \n
\n \n \n
Vorgestellte Projekte<\/h6>\n <\/div>\n <\/div>\n
\n \n
\n
\n \n \n \"\"\n
<\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n
\n
\n \n \n
\n \n
\n \n
\n \n \n
Kontakt<\/h6>\n <\/div>\n <\/div>\n
\n \n
\n \n
\n
\n \n \n
\n \n
\n \n
\n
\n \n \n \"\"\n
<\/div>\n <\/div>\n <\/div>\n <\/div>\n
\n \n
\n \n
\n
\n \n \n
\n \n
\n \n
\n \n \n
Frederik Schubert<\/h6>\n <\/div>\n <\/div>\n
\n \n
schubert@tnt.uni-hannover.de<\/span><\/a>\n <\/span>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n
\n \n \n
\n

Frederik Schubert ist wissenschaftlicher Mitarbeiter am L3S und am Institut f\u00fcr Informationsverarbeitung. Er besch\u00e4ftigt sich mit Reinforcement-Learning und der automatischen Szenenanalyse.<\/p>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n

\n
\n \n \n
\n \n
\n \n
\n
\n \n \n \"\"\n
<\/div>\n <\/div>\n <\/div>\n <\/div>\n
\n \n
\n \n
\n
\n \n \n
\n \n
\n \n
\n \n \n
Prof. Dr.-Ing. Bodo Rosenhahn<\/h6>\n <\/div>\n <\/div>\n
\n \n
rosenhahn@l3s.de<\/span><\/a>\n <\/span>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n
\n \n \n
\n

Bodo Rosenhahn ist Direktor des L3S und leitet das Institut f\u00fcr Informationsverarbeitung. Er forscht auf den Gebieten Computer Vision und Big Data.<\/p>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n

\n
\n \n \n
\n \n
\n \n
\n
\n \n \n \"\"\n
<\/div>\n <\/div>\n <\/div>\n <\/div>\n
\n \n
\n \n
\n
\n \n \n
\n \n
\n \n
\n \n \n
Amin Khoshandam<\/h6>\n <\/div>\n <\/div>\n
\n \n
amin.khoshandam@claimbuddy.de<\/span><\/a>\n <\/span>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n
\n \n \n
\n

Amin Khoshandam ist Technischer Direktor bei ClaimBuddy.<\/p>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n<\/div>","protected":false},"excerpt":{"rendered":"

Issue: 01\/2020 K\u00fcnstliche Intelligenz Automatic damage detection The roads in Germany are getting crowded. In the last ten years, the number of new car registrations has risen by over 30 percent. As the number of road users increases, so does the number of claims reported to insurance companies. Trained employees check the data and images […]<\/p>","protected":false},"author":7,"featured_media":7484,"parent":2750,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"page-templates\/full-width-page.php","meta":{"_acf_changed":false,"footnotes":""},"categories":[],"tags":[28,83],"class_list":["post-7483","page","type-page","status-publish","has-post-thumbnail","hentry","tag-binaire","tag-binaire-2020-01"],"acf":[],"aioseo_notices":[],"publishpress_future_action":{"enabled":false,"date":"2026-05-05 03:39:39","action":"category","newStatus":"draft","terms":[75],"taxonomy":"category","extraData":[]},"publishpress_future_workflow_manual_trigger":{"enabledWorkflows":[]},"_links":{"self":[{"href":"https:\/\/www.l3s.de\/de\/wp-json\/wp\/v2\/pages\/7483","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.l3s.de\/de\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/www.l3s.de\/de\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/www.l3s.de\/de\/wp-json\/wp\/v2\/users\/7"}],"replies":[{"embeddable":true,"href":"https:\/\/www.l3s.de\/de\/wp-json\/wp\/v2\/comments?post=7483"}],"version-history":[{"count":0,"href":"https:\/\/www.l3s.de\/de\/wp-json\/wp\/v2\/pages\/7483\/revisions"}],"up":[{"embeddable":true,"href":"https:\/\/www.l3s.de\/de\/wp-json\/wp\/v2\/pages\/2750"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.l3s.de\/de\/wp-json\/wp\/v2\/media\/7484"}],"wp:attachment":[{"href":"https:\/\/www.l3s.de\/de\/wp-json\/wp\/v2\/media?parent=7483"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.l3s.de\/de\/wp-json\/wp\/v2\/categories?post=7483"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.l3s.de\/de\/wp-json\/wp\/v2\/tags?post=7483"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}