Deep Hybrid Network for Land Cover Semantic Segmentation in High-Spatial Resolution Satellite Images

Khan, Sultan Daud; Alarabi, Louai; Basalamah, Saleh

doi:10.3390/info12060230

Cited by 30 publications

(15 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The method used is still static, so the use of dynamic video data cannot be done. The segmentation process can also use the deep feature [16]. This deep feature is suitable for highresolution images such as satellite images.…”

Section: Related Workmentioning

confidence: 99%

Real Time Fire Detection using Color Probability Segmentation and DenseNet Model for Classifier

Adhinata¹,

Ramadhan²

2022

IJACSA

View full text Add to dashboard Cite

The forest is an outdoor environment not touched by the surrounding community, so it is not immediately handled when a fire occurs. Therefore, surveillance using cameras is needed to see the presence of fire hotspots in the forest. This study aims to detect hotspots through video data. As is known, fire has a variety of colors, ranging from yellow to reddish. The segmentation process requires a method that can recognize various fire colors to get a candidate fire object area in the video frame. The methods used for the color segmentation process are Gaussian Mixture Model (GMM) and Expectation-maximization (EM). The segmentation results are candidates for fire areas, which in the experiment used the value of K=4. This fire object candidate needs to be ascertained whether the segmented object is a fire object or another object. In the feature extraction stage, this research uses the DenseNet-169 or DenseNet-201 models. In this study, various color tests were carried out, namely RGB, HSV, and YCbCr. The test results show that RGB color produces the most optimal training accuracy. This RGB color configuration is used to test using video data. The test results show that the true positive and false negative values are quite good, 98.69% and 1.305%. This video data processing produces fps with an average of 14.43. So, it can be said that this combination of methods can be used to process real time data in case studies of fire detection.

show abstract

Section: Related Workmentioning

confidence: 99%

Real Time Fire Detection using Color Probability Segmentation and DenseNet Model for Classifier

Adhinata¹,

Ramadhan²

2022

IJACSA

View full text Add to dashboard Cite

show abstract

“…They used the dense network in the encoder side to obtain global multi-scale deep features, whereas spatial information is recovered by using decoder. Authors in [16] presented a deep hybrid network for land cover semantic segmentation in high-spatial resolution satellite images. They combine two models, i.e., DenseNet and U-Net to perform pixel-wise classification of land cover.…”

Section: Related Workmentioning

confidence: 99%

Hybrid-DANet: An Encoder-Decoder Based Hybrid Weights Alignment With Multi-Dilated Attention Network for Automatic Brain Tumor Segmentation

et al. 2022

View full text Add to dashboard Cite

Gliomas are the most common and highly growing tumors lead to high mortality rate in their highest grade. The early diagnosis of gliomas, and treatment planning are most important steps to enhance the life expectancy of a patient. Among the modern imaging techniques, magnetic resonance imaging (MRI) is the most robust and widely used technique to visualize the brain tumor. The CNN-based networks mainly depend on multi-branch and increasing the depth/width of the network to enhance the segmentation accuracy at the cost of high computational cost. To mitigate these drawbacks we therefore, propose a hybrid weights alignment with multi-dilated attention network for automatic brain tumor segmentation (Hybrid-DANet). It employs multiple modules incorporated on baseline encoder-decoder architecture. Firstly, we proposed a novel hybrid weight alignment with multi-dilated attention module (HWADA) is used between the skip connections. It has capability to obtain the different sets of aligned weight by using different dilation schemes. Different weight alignments play a vital role to obtain very precise targeted information while negating the less informative part. It utilizes the low and high level information with skip connections across each branch of encoder and decoder. Secondly, we incorporated a multi channel multi scale module (MCS) on the baseline module. It consists of multiple channels used to extract the channel-wise information with more reduced computational cost. To reduce the resultant saturated accuracy due to vanishing gradient problem, we incorporated the residual module (RM). Thus, the RM, and MCS are useful to obtain the deep, intrinsic, channel-wise feature without expansion of depth and height. Whereas the novel HWADA not only propagates the low level information but also process it to obtain more semantic features used for decoder. We have tested our proposed technique on well-known datasets; BraTS 2017, and 2018 with comparable performance to counter-part.

show abstract

“…The authors in [14] and [15] use joint color texture representation called Hybrid Color Local Binary Patterns (HCLBP) and four supervised machine learning classification algorithms for texture analysis and texture classification. Some of the very recent work in this problem area is [16] based on deep learning.…”

Section: Related Workmentioning

confidence: 99%

A Unified Approach for Representing Outdoor Scenes

2022

View full text Add to dashboard Cite

Generalized image understanding of natural outdoor images is fundamentally a bottom-up approach guided by the data and information received from the digital input images. The importance of intermediate image representation that aids generalized image understanding, inspired the proposed a unified approach for providing a new meaningful and objective representation of a given outdoor scene. The proposed unified framework provides a novel set of string and semantic codes, based on inferential statistical data obtained from segments characterization and blocks characterization of natural outdoor images. The framework acts as an analytical tool that can categorize the input images into four types based on the inferential statistical representation -1) Object-based image, 2) Heterogeneous color based texture image, 3) Homogeneous color based texture image, 4) Pure toned image. The proposed string and semantic code characterization is also rotation and scale variant. Such a representation is simple, unsupervised, computationally inexpensive and also places minimal demand on the storage. Later for a higher-level image analysis and understanding the proposed unified approach can provide focused interaction between the above intermediate representation and domain scene knowledge for object detection, classification, or recognition.

show abstract

Deep Hybrid Network for Land Cover Semantic Segmentation in High-Spatial Resolution Satellite Images

Cited by 30 publications

References 57 publications

Real Time Fire Detection using Color Probability Segmentation and DenseNet Model for Classifier

Real Time Fire Detection using Color Probability Segmentation and DenseNet Model for Classifier

Hybrid-DANet: An Encoder-Decoder Based Hybrid Weights Alignment With Multi-Dilated Attention Network for Automatic Brain Tumor Segmentation

A Unified Approach for Representing Outdoor Scenes

Contact Info

Product

Resources

About