Indoor versus outdoor scene recognition for navigation of a micro aerial vehicle using spatial color gist wavelet descriptors

Anitha Ganesan,Anbarasu Balasubramanian

doi:10.1186/s42492-019-0030-9

Anitha Ganesan, Anbarasu Balasubramanian

Open Access

https://doi.org/10.1186/s42492-019-0030-9

Copy DOI

Abstract

In the context of improved navigation for micro aerial vehicles, a new scene recognition visual descriptor, called spatial color gist wavelet descriptor (SCGWD), is proposed. SCGWD was developed by combining proposed Ohta color-GIST wavelet descriptors with census transform histogram (CENTRIST) spatial pyramid representation descriptors for categorizing indoor versus outdoor scenes. A binary and multiclass support vector machine (SVM) classifier with linear and non-linear kernels was used to classify indoor versus outdoor scenes and indoor scenes, respectively. In this paper, we have also discussed the feature extraction methodology of several, state-of-the-art visual descriptors, and four proposed visual descriptors (Ohta color-GIST descriptors, Ohta color-GIST wavelet descriptors, enhanced Ohta color histogram descriptors, and SCGWDs), in terms of experimental perspectives. The proposed enhanced Ohta color histogram descriptors, Ohta color-GIST descriptors, Ohta color-GIST wavelet descriptors, SCGWD, and state-of-the-art visual descriptors were evaluated, using the Indian Institute of Technology Madras Scene Classification Image Database two, an Indoor-Outdoor Dataset, and the Massachusetts Institute of Technology indoor scene classification dataset [(MIT)-67]. Experimental results showed that the indoor versus outdoor scene recognition algorithm, employing SVM with SCGWDs, produced the highest classification rates (CRs)—95.48% and 99.82% using radial basis function kernel (RBF) kernel and 95.29% and 99.45% using linear kernel for the IITM SCID2 and Indoor-Outdoor datasets, respectively. The lowest CRs—2.08% and 4.92%, respectively—were obtained when RBF and linear kernels were used with the MIT-67 dataset. In addition, higher CRs, precision, recall, and area under the receiver operating characteristic curve values were obtained for the proposed SCGWDs, in comparison with state-of-the-art visual descriptors.

Highlights

Classification of a scene as being indoors or outdoors is a challenging task in the navigation of a micro aerial vehicle (MAV)
spatial color gist wavelet descriptor (SCGWD) are extracted from the test image, and the support vector machine (SVM) classifier with linear and nonlinear kernel classify the scene category as indoor or outdoor based on the scene recognition model learned at the training stage
area under the curve (AUC) Area under the receiver operating characteristic curve, CR Classification rate, radial basis function kernel (RBF) Radial basis function kernel, scale-invariant feature transform (SIFT)-LLC SIFT with locality-constrained linear coding, SIFT-Sparse coding based spatial pyramid matching (ScSPM) SIFT with sparse coding based spatial pyramid matching, SIFT-SPM SIFT with spatial pyramid matching, SPM Spatial pyramid matching, histogram of oriented gradients (HOG) Histogram of oriented gradients, speeded-up Robust Features (SURF) Speeded up robust features, CENTRIST Census transform histogram input image into 31 blocks, and the obtained histogram of Census Transformed values were concatenated into 31 blocks, to produce a 7936-dimensional descriptor

Summary

Introduction

Classification of a scene as being indoors or outdoors is a challenging task in the navigation of a micro aerial vehicle (MAV). Enhanced-GIST descriptor The Enhanced-GIST descriptor method was proposed [27] to recognize corridors, staircases, and room types, in indoor scenes, by encoding the spatial envelope and geometric structure of the indoor scene In this method, a bank of 32 Gabor filters is applied to an indoor and outdoor grayscale image (256 × 256 pixels), at 4 scales and 8 orientations, to produce 32 feature maps. A bank of 32 Gabor filters is applied to an indoor and outdoor grayscale image (256 × 256 pixels), at 4 scales and 8 orientations, to produce 32 feature maps These 32 feature maps are divided into 4 × 4 grids, and the filtered outputs within each of the 16 regions are averaged, to produce a 512-dimensional (16 × 32) GIST descriptor (Fig. 2g).

Methods

Conclusions