Fast detection of bag-breakups in pulsating and steady airflow using video analysis and deep learning

作     者：Morita, Daiki Raytchev, Bisser Elhanashi, Abdussalam Kawaguchi, Mikimasa Ogata, Yoichi Higaki, Toru Kaneda, Kazufumi Nakashima, Akira Saponara, Sergio 

作者机构：Hiroshima Univ Grad Sch Adv Sci & Engn Higashihiroshima Japan Univ Pisa Dip Ingn Informaz Pisa Italy Mazda Motor Corp MBD Innovat Dept Fuchu Japan 

出 版 物：《JOURNAL OF REAL-TIME IMAGE PROCESSING》 (J. Real-Time Image Process.)

年 卷 期：2023年第20卷第6期

页      面：114页

核心收录：

学科分类：0808[工学-电气工程] 1002[医学-临床医学] 08[工学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：This research was supported by Mazda Motor Corporation. The droplet dispersal related experiments from which the images were generated were carried out at the Mechanical Power and Motor Systems Laboratory, Hiroshima University. B.R. is supported in part by Mazda Motor Corporation [JP23K11170] JSPS KAKENHI 

主　　题：Object detection Scientific and industrial applications Real-time processing Small-size datasets YOLOv7 RetinaNet 

摘      要：Object detection methods based on deep learning have made great progress in recent years and have been used successfully in many different applications. However, since they have been evaluated predominantly on datasets of natural images, it is still unclear how accurate and effective they can be if used in special domain applications, for example in scientific, industrial, etc. images, where the properties of the images are very different from those taken in natural scenes. In this study, we illustrate the challenges one needs to face in such a setting on a concrete practical application, involving the detection of a particular fluid phenomenon-bag-breakup-in images of droplet scattering, which differ significantly from natural images. Using two technologically mature and state-of-the-art object detection methods, RetinaNet and YOLOv7, we discuss what strategies need to be considered in this problem setting, and perform both quantitative and qualitative evaluations to study their effects. Additionally, we also propose a new method to further improve accuracy of detection by utilizing information from several consecutive frames. We hope that the practical insights gained in this study can be of use to other researchers and practitioners when targeting applications where the images differ greatly from natural images.