Fully Convolutional Siamese Networks Based Change Detection for Optical Aerial Images with Focal Contrastive Loss

During the last 20 years, airborne laser scanning (ALS), often combined with multispectral information from aerial images, has shown its high feasibility for automated mapping processes. Recently, the first multispectral airborne laser scanners have been launched, and multispectral information is for the first time directly available for 3D ALS point clouds. This article discusses the potential of this new single-sensor technology in map updating, especially in automated object detection and change detection. For our study, Optech Titan multispectral ALS data over a suburban area in Finland were acquired. Results from a random forests analysis suggest that the multispectral intensity information is useful for land cover classification, also when considering ground surface objects and classes, such as roads. An out-of-bag estimate for classification error was about 3% for separating classes asphalt, gravel, rocky areas and low vegetation from each other. For buildings and trees, it was under 1%. According to feature importance analyses, multispectral features based on several channels were more useful that those based on one channel. Automatic change detection utilizing the new multispectral ALS data, an old digital surface model (DSM) and old building vectors was also demonstrated. Overall, our first analyses suggest that the new data are very promising for further increasing the automation level in mapping. The multispectral ALS technology is independent of external illumination conditions, and intensity images produced from the data do not include shadows. These are significant advantages when the development of automated classification and change detection procedures is considered.

Download Full-text

BUILDING CHANGE DETECTION FROM BITEMPORAL AERIAL IMAGES USING DEEP LEARNING

ISPRS Annals of Photogrammetry Remote Sensing and Spatial Information Sciences ◽

10.5194/isprs-annals-v-2-2020-565-2020 ◽

2020 ◽

Vol V-2-2020 ◽

pp. 565-571

Author(s):

S. Su ◽

T. Nawata ◽

T. Fuse

Keyword(s):

Deep Learning ◽

Change Detection ◽

Urban Areas ◽

Network Architecture ◽

Aerial Images ◽

Detection Accuracy ◽

Detection Model ◽

Wide Range ◽

Surface Models ◽

Data Source

Abstract. Automatic building change detection has become a topical issue owing to its wide range of applications, such as updating building maps. However, accurate building change detection remains challenging, particularly in urban areas. Thus far, there has been limited research on the use of the outdated building map (the building map before the update, referred to herein as the old-map) to increase the accuracy of building change detection. This paper presents a novel deep-learning-based method for building change detection using bitemporal aerial images containing RGB bands, bitemporal digital surface models (DSMs), and an old-map. The aerial images have two types of spatial resolutions, 12.5 cm or 16 cm, and the cell size of the DSMs is 50 cm × 50 cm. The bitemporal aerial images, the height variations calculated using the differences between the bitemporal DSMs, and the old-map were fed into a network architecture to build an automatic building change detection model. The performance of the model was quantitatively and qualitatively evaluated for an urban area that covered approximately 10 km2 and contained over 21,000 buildings. The results indicate that it can detect the building changes with optimum accuracy as compared to other methods that use inputs such as i) bitemporal aerial images only, ii) bitemporal aerial images and bitemporal DSMs, and iii) bitemporal aerial images and an old-map. The proposed method achieved recall rates of 89.3%, 88.8%, and 99.5% for new, demolished, and other buildings, respectively. The results also demonstrate that the old-map is an effective data source for increasing building change detection accuracy.

Download Full-text