An Improved InceptionV3 Network for Obscured Ship Classification in Remote Sensing Images

Abstract

Ship target classification plays an important role in tasks such as maritime traffic control, maritime target tracking, and military reconnaissance. The complex ocean environment often causes obscuration of the ship targets, thus resulting in low accuracy of the obscured targets. This paper presents a novel target classification algorithm-Improved InceptionV3 and Center Loss CNN (IICL-CNN)-based on the well-established Inception network to improve the accuracy of obscured targets. This algorithm features a new objective function, which is designed to learn common features of both the clear samples and the obscured samples and, in the meantime, reduce the intra-class distance among the obscured samples. Experiments were performed on an optical remote sensing image dataset which consisted of 48,000 ship images in 9 categories. The proposed method demonstrated superior performance on the obscured ship targets compared to the original InceptionV3 model. On average, the accuracy was 4.23%, 5.98%, and 17.48% higher on the ship targets that were occluded by levels of 30%, 50%, and 70%, respectively. Our experimental results showed that the proposed IICL-CNN could effectively improve the accuracy of the ship targets at various occlusion levels. Index Terms-Convolution neural network, deep learning, feature extraction, fog occlusion, image classification, remote sensing image I. INTRODUCTION HIP target classification is an emerging research area which has a broad range of applications in civil and military fields, such as maritime traffic control, onshore search, maritime target tracking, and military reconnaissance. With the developments in computer vision and machine learning, there are many recent studies in this area. Zhu et al. [1] recently proposed a method for ship identification, which extracted multiple highdimensional local features from ship images and then used Support Vector Machine (SVM) for target classification. In order to identify ship targets more accurately, SVM is combined with other methods to improve effectiveness and robustness [2, 3, 4, 5]. Aiming at the sparseness of targets in optical remote sensing images, Li et al. [6] proposed a method for target detection and classification based on morphological

3DeepM: An Ad Hoc Architecture Based on Deep Learning Methods for Multispectral Image Classification

“…The architectures reviewed in this work used in remote sensing applications have been specifically designed to be used with hyper-or multispectral images. In most cases [40][41][42][43][44][45][47][48][49], the authors apply the transfer learning technique, which means that they used a model that was not specifically trained for their case of study and adapted it via fine-tuning.…”

“…VGG16 and 19, InceptionV3, Xception and ResNet50, among others, are used in a comparative study [44] to classify complex multispectral optical imagery of wetlands. In [45], the authors designed an improved version of InceptionV3 to classify ship imagery from optical remote sensors. This architecture is also the base of the one used in [46] to classify images of damaged buildings by earthquakes using highresolution remote sensing images; and in [47], the authors pretrained an ImageNet2015 InceptionV3 model together with VGG19 and ResNet50 to classify images of high spatial resolution.…”

Comparative Study of Ensemble Models of Deep Convolutional Neural Networks for Crop Pests Classification

“…InceptionV3 is an implementation of GoogLeNet, its ability to deconstruct these features into smaller convolution sections (Zhao et al 2020). Its network can be efficiently decomposed into small convolution kernels, which greatly reduces the number of parameters of the model and the chance of overfitting (Liu et al 2020). This paper freeze the first 270 layers of InceptionV3 and add three fully connected layers containing 512 neurons.…”

CenterNet-SPP based on multi-feature fusion for basketball posture recognition

“…Inception V3 network is a deep convolutional network developed by Google [13]. The main idea of Inception structure in the model is to find out how to approximate the optimal local sparse structure with dense components.…”