Region-based Segmentation of Terrain Fields in SAR Images

Yayla, Rıdvan; Şen, Baha

doi:10.1109/siu49456.2020.9302432

Cited by 5 publications

(6 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The general structure of Mask R-CNN is given in figure 4. In Mask R-CNN, the loss function (L) is the sum of classification losses (Lcls), masking losses (Lmask) and frame losses (Lbox) [23]. The loss function is shown in Equation ( 1).…”

Section: Methodsmentioning

confidence: 99%

Military camouflage classification with Mask R-CNN algorithm

KARATEPE,

NABİYEV

2023

Communications Faculty of Sciences University of Ankara Series A2-A3 Physical Sciences and Engineering

View full text Add to dashboard Cite

Camouflage, which is used as an art of hiding by living things in nature, started to be used in the military field in the 19th century with the widespread use of long-range firearms. When factors such as different nations, environment and climate are considered, we come across camouflages in various colors and patterns. Over time, the camouflage patterns adopted and used by countries or unions have become national identity. This study is on the classification and segmentation of camouflaged soldiers of 5 countries with deep learning. While the similarity of the camouflaged area with the background makes segmentation difficult, it becomes difficult to classify each camouflage pattern due to the cut of the fabric and the different locations of the pattern pieces on each soldier. There are different studies in the literature that are referred to as camouflage or pattern classification. The mentioned studies are in the form of segmentation of camouflaged object or classification of camouflaged objects of different types. Since the segmented and classified objects in this study are camouflaged soldiers, what is expected from the deep learning algorithm is to classify the objects mainly according to the camouflage pattern, not their outlines. In the study, 861 camouflaged soldier images were collected for 5 countries (Türkiye-Azerbaijan, USA, Russia, China, France) and polygonal labeling was made. Türkiye and Azerbaijan are considered a class as they have similar camouflages. For the solution of the problem, military camouflage classification was discussed with the Mask R-CNN algorithm, which is widely used today for object detection, segmentation and classification, and the importance of deep learning algorithms has been proven with such a difficult problem. The training resulted in 0.005219 classification loss and 0.03985 masking loss. The classification and segmentation success rate of the study is 95%.

show abstract

Section: Methodsmentioning

confidence: 99%

Military camouflage classification with Mask R-CNN algorithm

KARATEPE,

NABİYEV

2023

Communications Faculty of Sciences University of Ankara Series A2-A3 Physical Sciences and Engineering

View full text Add to dashboard Cite

show abstract

“…There are many kinds of classical image segmentation methods for sea clutter, e.g., threshold‐based method Xu et al (2010), region‐based method Yayla and Şen (2020), edge‐based method M. Wang et al (2008), etc. Among them, the threshold‐based methods (e.g., grey histogram‐based method, Otsu‐based method, Iteration‐based method, Entropy histogram‐based method, etc.)…”

Section: Related Workmentioning

confidence: 99%

Marine radar monitoring IoT system and case study of target detection based on PPI images

Zhang

Guo

et al. 2023

Expert Systems

View full text Add to dashboard Cite

To ensure the ship navigation safety, it is necessary to detect the targets in the background of sea clutter around the ship. The most commonly used target detection equipments for ship navigation are marine radar and automatic identification system (AIS) device. However, marine radar echoes are often mixed with multiple clutter, and weak targets are not always equipped with AIS device, neither provides AIS information with low accuracy, leading to difficulty with target detection. Moreover, existing marine monitoring systems are accustomed to using the traditional information fusion methods for target detection, and the accuracy of identifying the weak targets is relatively low. To make full use of radar echo and AIS information and improve the accuracy of target detection, a Marine radar monitoring Internet of Things system is proposed, in which marine radars work in both scanning and staring modes. By adopting image segmentation and deep learning methods, the proposed design can enable accurate perception of weak target detection based on plan‐position indicator (PPI) images. A case study based on the selected X‐band radar datasets shows that the proposed design can achieve high target identification accuracy.

show abstract

“…In equation ( 3), L cls expresses the classification loss, L box is the bounding box loss, and L mask is the masking loss. RoIPool is a useful process for exposing a small feature map such as 7×7 from each RoI [17]. RoI Align abolishes the harsh quantization of RoI Pool by aligning the determined features with the input.…”

Section: Mask R-cnnmentioning

confidence: 99%

“…RoI Align abolishes the harsh quantization of RoI Pool by aligning the determined features with the input. The bilinear interpolation which computes the exact values of the input features at four regularly sampled locations in each RoI bin and aggregates the result (using max or average) is used in this model [16] [17]. The Mask R-CNN differs from the Faster R-CNN with two features.…”

Section: Mask R-cnnmentioning

confidence: 99%

See 1 more Smart Citation

Vehicle Detection from Unmanned Aerial Images with Deep Mask R-CNN

Yayla¹,

Albayrak²,

Yüzgeç³

2022

CSJM

View full text Add to dashboard Cite

In this paper, a classification approach which is applied to Mask Region-based Convolutional Neural Network as deeper is proposed for vehicle detection on the images from UAV instead of the familiar methods. The different types of unmanned aerial vehicles are widely used for a lot of areas such as agricultural spraying, advertisement shooting, fire extinguishing, transportation and surveillance, exploration, destruction for the military. In recent years, deep learning techniques are progressively developed for object detection. Segmentation algorithms based on CNN architecture are especially widely used for extracting meaningful parts of an image. Additionally, Mask R-CNN based on CNN architecture rapidly detects the object with high-accuracy on an image. This study shows that the high-accuracy results are obtained when the Mask R-CNN is applied as deeper in vehicle detection on the images taken by UAV.

show abstract

Region-based Segmentation of Terrain Fields in SAR Images

Cited by 5 publications

References 6 publications

Military camouflage classification with Mask R-CNN algorithm

Military camouflage classification with Mask R-CNN algorithm

Marine radar monitoring IoT system and case study of target detection based on PPI images

Vehicle Detection from Unmanned Aerial Images with Deep Mask R-CNN

Contact Info

Product

Resources

About