Research for detection and flow analysis of moving objects with shape change and occlusion

2019 Fiscal Year Final Research Report

• Project/Area Number: 16K06387
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Measurement engineering
• Research Institution: Tottori University
• Principal Investigator: KONDO Katsuya 鳥取大学, 工学研究科, 教授 (00295750)
• Co-Investigator(Kenkyū-buntansha): 三柴 数 鳥取大学, 工学研究科, 准教授 (40609038) ; 小山田 雄仁 鳥取大学, 工学研究科, 助教 (30708615) ; 栗政 明弘 東北医科薬科大学, 医学部, 教授 (80343276)
• Project Period (FY): 2016-04-01 – 2020-03-31
• Keywords: 計測工学 / 画像解析 / 細胞

Outline of Final Research Achievements

In microscopic cell images, cell brightness changes due to cell position in depth, so there are cells with various brightness mixed in an image. Then it tends to miss the detection and tracking of cells, and it is difficult to evaluate the growth and genealogy analysis. By suppressing undetected cells for various brightness ranges, we solved this subject so that we show that it works well to improve the detection for the accuracy of flow and genealogy imaging. For person detection problem in the image taken from a floating camera, we reduced the effect of partial hiding by swaying trees. Furthermore, by using temperature features, we showed that it is reidentificated after occluded.

Free Research Field

画像計測

Academic Significance and Societal Importance of the Research Achievements

再生医療の高まりのもと幹細胞の成長過程の評価，群れた人の動きの解析は種々の応用が期待できる．細胞や人などの動きを解析することは，精度よく個々を分離して動きを追跡することが必要で，実用上，近接・接触や隠蔽など特有の問題が存在し，画像内の個々の動き評価を難しくしている．本課題はいくつかの解決法を示し，その適用範囲，応用範囲を増やすことに寄与するものである．また近年，熱画像を用いた動線解析が盛んになってきており，熱特徴に基づく人物追跡も扱った．