logo

EbookBell.com

Most ebook files are in PDF format, so you can easily read them using various software such as Foxit Reader or directly on the Google Chrome browser.
Some ebook files are released by publishers in other formats such as .awz, .mobi, .epub, .fb2, etc. You may need to install specific software to read these formats on mobile/PC, such as Calibre.

Please read the tutorial at this link:  https://ebookbell.com/faq 


We offer FREE conversion to the popular formats you request; however, this may take some time. Therefore, right after payment, please email us, and we will try to provide the service as quickly as possible.


For some exceptional file formats or broken links (if any), please refrain from opening any disputes. Instead, email us first, and we will try to assist within a maximum of 6 hours.

EbookBell Team

Deep Learning Based Vehicle Detection In Aerial Imagery 1st Lars Wilko Sommer

  • SKU: BELL-230260508
Deep Learning Based Vehicle Detection In Aerial Imagery 1st Lars Wilko Sommer
$ 31.00 $ 45.00 (-31%)

4.1

60 reviews

Deep Learning Based Vehicle Detection In Aerial Imagery 1st Lars Wilko Sommer instant download after payment.

Publisher: KIT Scientific Publishing
File Extension: PDF
File size: 4.53 MB
Pages: 276
Author: Lars Wilko Sommer
ISBN: 9783731511137, 3731511134
Language: English
Year: 2022
Edition: 1st
Volume: 52

Product desciption

Deep Learning Based Vehicle Detection In Aerial Imagery 1st Lars Wilko Sommer by Lars Wilko Sommer 9783731511137, 3731511134 instant download after payment.

Object detection in aerial imagery is essential for a wide range of applications in the field of civil safety and security. However, low spatial resolution originating from the large distance between sensor and ground, capturing conditions and varying scenarios due to different daytimes and regions impede the detection task.  This book proposes a novel deep learning based detection method, focusing on vehicle detection in aerial imagery recorded in top view. The base detection framework is extended by two novel components to improve the detection accuracy by enhancing the contextual and semantical content of the employed feature representation, yielding a reduced number of false detections. To reduce the computational effort and consequently the inference time, a lightweight CNN architecture optimized with regard to vehicle detection in aerial imagery is proposed as base architecture and a novel module restricting the search area to areas of interest is introduced. Extensive evaluation demonstrates state-of-the-art performance and the generalization ability on unseen aerial imagery data.

Related Products