Abstract
In remote sensing, active learning (AL) is considered to be an effective solution to the problem of producing sufficient classification accuracy with a limited number of training samples. Though this field has been extensively studied, most papers exist in the pixel-based paradigm. In object-based image analysis (OBIA), AL has been comparatively less studied. This paper aims to propose a new AL method for selecting object-based samples. The proposed AL method solves the problem of how to identify the most informative segment-samples so that classification performance can be optimized. The advantage of this algorithm is that informativeness can be estimated by using various object-based features. The new approach has three key steps. First, a series of one-against-one binary random forest (RF) classifiers are initialized by using a small initial training set. This strategy allows for the estimation of the classification uncertainty in great detail. Second, each tested sample is processed by using the binary RFs, and a classification uncertainty value that can reflect informativeness is derived. Third, the samples with high uncertainty values are selected and then labeled by a supervisor. They are subsequently added into the training set, based on which the binary RFs are re-trained for the next iteration. The whole procedure is iterated until a stopping criterion is met. To validate the proposed method, three pairs of multi-spectral remote sensing images with different landscape patterns were used in this experiment. The results indicate that the proposed method can outperform other state-of-the-art AL methods. To be more specific, the highest overall accuracies for the three datasets were all obtained by using the proposed AL method, and the values were 88.32%, 85.77%, and 93.12% for “T1,” “T2,” and “T3,” respectively. Furthermore, since object-based features have a serious impact on the performance of AL, eight combinations of four feature types are investigated. The results show that the best feature combination is different for the three datasets due to the variation of the feature separability.
Highlights
The number of remote sensing platforms is continually increasing, and they are producing a tremendous amount of earth-observation image data
An object-based active learning algorithm has been proposed in this article
The objective of this method is to select the most useful segment-samples, which are to be labeled by a supervisor and added to the training set to improve classification accuracy
Summary
The number of remote sensing platforms is continually increasing, and they are producing a tremendous amount of earth-observation image data. Image classification plays a fundamental role in this task, and there have been great efforts dedicated to the development of image classification methods, such as data mining, deep learning [4,5], and object-based image analysis (OBIA) Among these techniques, OBIA has been widely applied to high spatial resolution image classification [6,7,8] because it can take high advantage of the spatial information that is captured within these images. OBIA-based classification algorithms can reduce or even eliminate salt-and-pepper noises, which often exist in the classification results of pixel-based strategies [18] This is because image segmentation is generally the first step in OBIA. Driven by Tobler’s first law of geography, Lv et al [23] extracted object-based spatial–spectral features to enhance the classification accuracy of aerial imagery
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
