T. Chanwimaluang - IEEE Xplore Author Profile

Showing 1-25 of 40 results

Filter Results

Show

Results

In a supervised classification of a remote sensing image, a ground survey is needed to collect sufficient training data to train a classifier. The ground survey is expensive and time consuming. Thus, the idea of the class activation map (CAM) is proposed for land cover mapping in this paper. Here, instead of training with ground data where the underlying class label of each training pixels must be...Show More
This paper represents an application of a Deep Qlearning in the rice crop cultivation practice, where the optimal actions are determined. In this work, we focused on rain-fed rice where the main actions are when to start cultivation and when to harvest. The goal is to find the optimum cultivation and harvest period such that farmers' income is maximized. To make our study realistic, the actual cli...Show More
The 2-D Forward-looking Sonar (2D-FLS) is a new technology of sonar signals that increases the performance of underwater exploration of an Autonomous Underwater Vehicle (AUV). Because the AUV is a fully automatic robot that can handle various missions and does not require humans to control its operation, FLS is the best tool for underwater navigation in poor visibility conditions. In this paper, w...Show More
Autonomous underwater vehicles (AUVs), unmanned underwater vehicles, provide lower operation costs in underwater pipeline inspection than remotely operated vehicles (ROVs), since there is no need for a human operator. However, the navigation algorithm is an important part of the successful deployment of AUVs. In this work, we propose an algorithm that extracts the direction of a pipeline based on ...Show More
Autonomous underwater vehicles (AUVs), unmanned underwater vehicles, are cheaper to operate than remotely operated vehicles (ROVs), since an AUV does not require a trained human operator to control. As a result, in this paper, we propose the pipeline tracking algorithm using sequences of the forward-looking sonar (FLS) images installed on a preprogram AUV. The FLS is more suitable to be used in th...Show More
Traditionally, the land cover mapping process needs a ground data to be collected with high precision in both class labeling and spatial locations. To collect enough, high precise ground data require resources. As a result, we proposed an approach for building an image classification based on the class activation map (CAM) where the goal is not to identify the relationship between each pixel and a...Show More
The underwater pipeline surveillance and inspection using an autonomous underwater vehicle (AUV) is the future of the oil and gas industries since the inspection using a remotely operated underwater vehicle (ROV) is very expensive. However, the successful deployment of AUVs relies on the accurate pipeline path extraction that can automatically navigate along its direction in real-time. Due to the ...Show More
The uncontrollable weather conditions can cause a serious problem to remote sensing imaginary. One of the weather conditions is a resulting from cloud contamination. As a result, this paper proposed the use of the convolutional autoencoder neural networks to remove clouds from cloud-contaminated images by training on a multi-temporal remote sensing dataset. Here, the observations from different sp...Show More
In this paper, we proposed an approach for super-resolution land cover mapping on remote sensing images based on the deep learning technique, namely Convolutional Neural Network (CNN) by combining with the level set method (LSM). Here, the CNN is used to find the probabilities that a subpixel belonging to a land cover class, and the LSM is employed to fine tune the boundaries among land cover clas...Show More
A mixed pixel in remote sensed images is a major problem, and the super-resolution mapping is one of the approach to deal with this problem. In this paper, we address the problem of super-resolution mapping by combining a set of random forests with a Markov random field (MRF) model. Here, a random forest is trained to estimate a class proportion of only one land cover class. Thus, there are equal ...Show More
Now a day's recognition of satellite image authenticity has received too much attention due to the invention of various remote sensing image inpainting algorithms. Satellite image forgery can be referred as a technique in which fake satellite image is generated by the creation and alternation of new image contents. This paper proposes an algorithm for the identification of inpainted remote sensing...Show More
Image inpainting refers to a technique in which missing areas of an image are filled is such a way that it looks plausible to the human eye by retrieving information from the surrounding pixels. Degradation in remote sensing images is usually caused by dead pixels, noise, clouds, sensor problem or communication system problem. The aim of this paper is to evaluate the performance of different inpai...Show More
This paper proposes a new land cover mapping algorithm that combines the strengths of random forest (RF) with a Markov random field (MRF) model. The idea is to transform the observed data into the decision domain of weak classifiers inside an RF. Due to how RF are trained, these decisions can be considered to be independent from each others, and therefore the joint probability density function in ...Show More
In this paper, we introduced a land cover mapping algorithm that combines for unsupervised and supervised classification techniques, namely, the Restricted Boltzmann machines (RBMs) and Support Vector Machines (SVMs). The idea is to take advantage of unsupervised classifications that can segment an image into regions without any training samples, and the supervised classification that can identify...Show More
In this paper, we introduce a new land cover mapping technique by taking advantages of a weighted random forest [1] and the level set method [2] to remove the weaknesses of each other. The weighted random forest can accurately estimate the likelihood that a pixel belonging to each classes while the level set method can capture the dependency among neighboring pixels. As a result, by combining thei...Show More
In this paper, we proposed a new random forest algorithm designed specifically for the land cover mapping problem. Three approaches are investigated, namely, pixel-based, neighbor-looking and combination of both. In the pixel-based approach, we use the fact that all decision trees are different whereas, in the neighbor-looking, the decisions from neighboring pixels are used when the decisions from...Show More
In this paper, we propose a level-set based method to identify urban areas using multitemporal Synthetic Aperture Radar (SAR) data. Our method is compared to the standard method, called the Otsu's threshold-based method. The experimental results indicate significant improvement in terms of the Kappa coefficients and the percentage of correctly classified pixels.Show More
We have proposed a new speckle filtering method in this paper. Our method uses time-series of SAR images at the same scene to perform multitemporal speckle filtering to reduce speckle noise while preserving its spatial information. We used segmentation based on the distribution of speckle noise by the Öztürk algorithm in this method for achieve a higher quality of edge areas.Show More
The actual field survey data from the Rice Department of Thailand's Ministry of Agriculture over a large area wastes a huge amount of resources. To solve this problem, this paper proposes a new approach to estimate rice phenology using SAR images derived from the RADARSAT-2 data. In this work, we divided the rice phenology into five stages, consisting of seedling, tillering, reproductive, ripening...Show More
In this paper, we propose the speckle removal algorithm from Synthetic Aperture Radar (SAR) images via a segmentation technique based on Öztürk algorithm. The Öztürk's algorithm is used to determine the distribution for each sample. These procedures are processed to see the relation between distribution of each pixels and segments of the image. We demonstrated the effectiveness of our algorithm by...Show More
The actual field survey data from the Rice Department of Thailand's Ministry of Agriculture over a large area wastes a huge amount of resources. To solve this problem, this paper proposes a new approach to estimate rice phenology using SAR images derived from the RADARSAT-2 data. In this work, we divided the rice phenology into five stages, consisting of seedling, tillering, reproductive, ripening...Show More
In this paper, we propose a level-set method to identify urban areas using a nighttime light data of Suomi National Polar-orbiting Partnership (NPP) Visible Infrared Imaging Radiometer Suite (VIIRS). Our method is compared to two standard methods, called the Otsu's threshold-based method and the k-mean clustering method. The experimental results indicates significant improvement in terms of the Ka...Show More
This paper addresses the problem of flood detection from the cloud-contaminated MODIS time-series data. Although MODIS data can provide almost daily coverage over the large area with the medium resolution. The use of MODIS data for flood mapping in the tropical regions is a challenging task due to the cloud contamination. Since the floods usually occur in the connected regions over a certain perio...Show More
In this paper, we proposed an unsupervised algorithm to identify the flooded areas from synthetic aperture radar (SAR) images based on texture information derived from the gray-level co-occurrence matrices (GLCM) texture analysis. Here, five GLCM features, namely, energy, contrast, homogeneity, correlation and entropy, are extracted from a SAR image. These features are input to an image segmentati...Show More