Sparse Pixel Training of Convolutional Neural Networks for Land Cover Classification

Abstract

Convolutional Neural Networks (CNN) have become the core of modern machine learning approaches. In addition to its inspiring interior design idea, the success of CNN depends mainly on two factors, the first is the availability of training data and the second is the computing power of the used devices. In the field of remote sensing, data availability is difficult and expensive. Furthermore, processing large remote sensing data to accommodate different models is a laborious process. At the same time, training data is often collected in the form of points distributed over crop fields rather than regions, which results in the scarcity of training data. To specifically address the scarcity of training points, in this paper we present a sparse pixel-based training of U-Net convolutional neural networks for land cover classification. Training images are reconstructed from the points' collection in a random manner, they are used as an input for the convolution networks. Based on this proposed method, the amount of training data is reproduced from the different spectral signals for each land cover. We conducted extensive experiments on eight classes, using ground truth data collected from several locations in Fayoum Governorate, Egypt. The obtained results showed the superiority of the proposed method over other methods.

Land-Cover-Change Detection with Aerial Orthoimagery Using SegNet-Based Semantic Segmentation in Namyangju City, South Korea

“…Unlike aerial orthoimages, which typically use only the RGB spectrum, multispectral images present data from various wavelengths. Laban et al used 10 bands of data from Sentinel satellite images with a spatial resolution of 10 m, and they reported an OA of 71% for eight land-cover types [45]. Zhang et al used the RGB and NIR bands of ZY-3 images with a spatial resolution of 5.8 m, and they reported an OA of 79.4% for six land-cover types [46].…”

Shallow Parallel CNNs for contextual remote sensing image classification

“…KNN performs very well in many problems, especially in the case of a lack of available resources such as memory or processing power [10]. Usually, KNN works as a baseline method for comparison in classification operations [8,18]. It attributes a pixel a class after examining the K nearest training sample in the feature class.…”

“…Thereafter, these markers may be used as pairs of georeferenced location and class type for each training pixel in the remote sensing image. classification methods ingests training data either as individual labeled pixels or labeled polygons (a region of labeled pixels that composes one object) [8].…”