An Object-Based Image Analysis Method for Enhancing Classification of Land Covers Using Fully Convolutional Networks and Multi-View Images of Small Unmanned Aerial System

Abstract

Fully Convolutional Networks (FCN) has shown better performance than other classifiers like Random Forest (RF), Support Vector Machine (SVM) and patch-based Deep Convolutional Neural Network (DCNN), for object-based classification using orthoimage only in previous studies; however, for further improving deep learning algorithm performance, multi-view data should be considered for training data enrichment, which has not been investigated for FCN. The present study developed a novel OBIA classification using FCN and multi-view data extracted from small Unmanned Aerial System (UAS) for mapping landcovers. Specifically, this study proposed three methods to automatically generate multi-view training samples from orthoimage training datasets to conduct multi-view object-based classification using FCN, and compared their performances with each other and also with RF, SVM, and DCNN classifiers. The first method does not consider the object surrounding information, while the other two utilized object context information. We demonstrated that all the three versions of FCN multi-view object-based classification outperformed their counterparts utilizing orthoimage data only. Furthermore, the results also showed that when multi-view training samples were prepared with consideration of object surroundings, FCN trained with these samples gave much better accuracy than FCN classification trained without context information. Similar accuracies were achieved from the two methods utilizing object surrounding information, although sample preparation was conducted using two different ways. When comparing FCN with RF, SVM, DCNN implies that FCN generally produced better accuracy than the other classifiers, regardless of using orthoimage or multi-view data.Remote Sens. 2018, 10, 457 2 of 24 all the UAS images. Then, image segmentation algorithm is conducted to segment the orthoimage to groups of homogeneous pixels to form numerous meaningful objects. Spectral, geometrical, textural, and contextual features are extracted from these objects and used as input to different classifiers, such as Random Forest (RF) [9] and Support Vector Machine (SVM) [10], to label the objects. Feature extraction and selection that have to be conducted during traditional OBIA procedures are challenging tasks and can limit classification performance.Recently, the rise of deep learning techniques provided an alternative to traditional land cover classifiers. Deep learning brought about around 2006 [11], became well known in the computer vision community around 2012, since one supervised version of deep learning networks Deep Convolutional Neural Networks (DCNN) made a breakthrough for scene classification tasks [12,13], and has reached out to many industrial applications and other academic areas in recent years as it continues to advance technologies in areas, like speech recognition [14], medical diagnosis [15], autonomous driving [16], or even the gaming world [17,18]. When compared with other traditional classifiers, deep learning does not require feature e...

Land Cover Classification from fused DSM and UAV Images Using Convolutional Neural Networks

“…Images captured using UAVs are used for geographical information system databases, datasets for automated decision-making, agricultural mapping, urban planning, land use and land cover detection and environmental monitoring and assessment [1,[5][6][7]. Such images are commonly used in supervised machine learning-based classification tasks as training data [8][9][10]. One reason for this is that these images have high resolution and a good range of spectral bands [6].…”

Integration of Convolutional Neural Networks and Object-Based Post-Classification Refinement for Land Use and Land Cover Mapping with Optical and SAR Data

“…CNNs are originally designed for image recognition, and the input shall be a rectangular image [49][50][51]. Zhao et al [52] applied a five-layer CNN to extract spatial features within an 18 × 18 window.…”

Robinia pseudoacacia L. in Short Rotation Coppice: Seed and Stump Shoot Reproduction as well as UAS-based Spreading Analysis

“…The CNN applied in the current study was tested to classify single black locust images under varying conditions and attained a high test accuracy of 99.5%. However, disadvantages of CNN algorithms are that a high number of labeled training images should be available [146], as well as the difficult traceability of the used classification features [105]. Nevertheless, studies [73,145,146] have shown that when the training sample size was high, CNN tended to show better results and accuracies compared to random forest, support vector machine, and fully convolutional networks.…”

“…However, disadvantages of CNN algorithms are that a high number of labeled training images should be available [146], as well as the difficult traceability of the used classification features [105]. Nevertheless, studies [73,145,146] have shown that when the training sample size was high, CNN tended to show better results and accuracies compared to random forest, support vector machine, and fully convolutional networks. Therefore, Liu and Abd-Elrahman [146] used 400 UAS images per object, Diegues et al [147] applied about 700 underwater images, Abrams et al [148] operated with 700 canopy and 800 understory images per habitat class as well as Li et al [73] used 5,000 satellite images per category.…”