写真測量とリモートセンシング
Online ISSN : 1883-9061
Print ISSN : 0285-5844
ISSN-L : 0285-5844
58 巻 , 3 号
選択された号の論文の9件中1~9を表示しています
巻頭言
カメラアイ
小特集「空間情報に関する国際標準化活動」
原著論文
  • 朱 林, 島村 秀樹
    2019 年 58 巻 3 号 p. 108-122
    発行日: 2019年
    公開日: 2020/07/01
    ジャーナル フリー

    In this study, we have developed a novel methodology for building change detection in the dense urban areas. Our approach is based on building recognition using aerial images and Digital Surface Models (DSMs) that allows detection of large and small buildings respectively. Large buildings are detected by a global thresholding of the DSMs using a height threshold in order to prevent the problem that one building may be divided into several portions. Small gable-roof and flat-roof buildings are detected individually according to the three-dimensional shape of the roofs so that buildings can be separated from each other more easily in densely built-up areas. Afterwards, change detection is implemented based on the result of building recognition, and only the DSMs are used for detecting the change of buildings in order to avoid the influence of image color variation. Also, in order to detect partially-changed buildings accurately, the increased and decreased height differences of two epochs of DSMs are extracted individually, and image morphological processing is performed to remove noise and extract actual changed areas. To assess the effectiveness of the proposed methodology, the change detection result has been verified by comparing to a visual interpretation result. The experimental results indicate 78.1% completeness with correctness of 52.3% in a dense built-up area, which demonstrate that our methodology can stably detect changed buildings with variation in height such as newly constructed, demolished, extended and structural alterations, and suppress false detection effectively.

  • 前原 秀明, 長瀬 百代, 服部 亮史, 平 謙二
    2019 年 58 巻 3 号 p. 123-129
    発行日: 2019年
    公開日: 2020/07/01
    ジャーナル フリー

    We have been studying a imaging type water level measurement method using images recorded water gauges which look like rulers been stuck into rivers vertically. It is the feature of this unit that observed images are collated with the one prerecorded in the time of low water level in order to specify the water border position. We implemented a program based on our method and tested the performance using multiple actual video data acquired under different weather and sunshine conditions and so on. As the result, which shows that 24 hours continuous observation can be accomplished and the calculated value is satisfied within required accuracy at typical video data, we have made sure of the effectiveness of our method.

  • 李 勇鶴, 坂元 光輝, 篠原 崇之, 佐藤 俊明
    2019 年 58 巻 3 号 p. 130-141
    発行日: 2019年
    公開日: 2020/07/01
    ジャーナル フリー

    In recent years, extensive researches have been conducted to automatically generate high-resolution road orthophotos using images and laser point cloud data collected by a Mobile Mapping System (MMS). However, it is necessary to detect and mask out the areas of non-road objects in MMS images such as vehicles, bicycles, pedestrians and their shadows, in order to eliminate erroneous textures from the road orthophotos. Hence, we proposed a novel vehicle and its shadow detection method based on Faster R-CNN for improving the detection accuracy, especially the accuracy of detected regions. The experimental results showed that the recall of the proposed method was 93.9% (Intersection-over-Union>0.7), which was 7.0% higher than 86.9% obtained by Faster R-CNN. Moreover the proposed method could identify the regions of vehicles and their shadows accurately and robustly in MMS images, even though the images contained various types of vehicles, different shadow directions, and partial occlusions. Furthermore, it was confirmed that the quality of road orthophoto generated using vehicle and its shadow masks by the proposed method was significantly improved as compared to those generated using no masks, vehicle masks and even the vehicle and its shadow masks by Faster R-CNN.

feedback
Top