• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.3 no.1
• /
• pp.52-57
• /
• 2003

In this paper, we propose a Fuzzy Classifier System(FCS) to find a set of fuzzy rules which can carry out the edge detection. The classifier system of Holland can evaluate the usefulness of rules represented by classifiers with repeated learning. FCS makes the classifier system be able to carry out the mapping from continuous inputs to outputs. It is the FCS that applies the method of machine learning to the concept of fuzzy logic. It is that the antecedent and consequent of classifier is same as a fuzzy rule. In this paper, the FCS is the Michigan style. A single fuzzy if-then rule is coded as an individual. The average gray levels which each group of neighbor pixels has are represented into fuzzy set. Then a pixel is decided whether it is edge pixel or not using fuzzy if-then rules. Depending on the average of gray levels, a number of fuzzy rules can be activated, and each rules makes the output. These outputs are aggregated and defuzzified to take new gray value of the pixel. To evaluate this edge detection, we will compare the new gray level of a pixel with gray level obtained by the other edge detection method such as Sobel edge detection. This comparison provides a reinforcement signal for FCS which is reinforcement learning. Also the FCS employs the Genetic Algorithms to make new rules and modify rules when performance of the system needs to be improved.

• Journal of Information Technology Services
• /
• v.15 no.4
• /
• pp.85-95
• /
• 2016

Recently, deep learning and machine learning have attracted considerable attention and many supporting frameworks appeared. In artificial intelligence field, a large body of research is underway to apply the relevant knowledge for complex problem-solving, necessitating the application of various learning algorithms and training methods to artificial intelligence systems. In addition, there is a dearth of performance evaluation of decision making agents. The decision making agent that can find optimal solutions by using reinforcement learning methods designed through this research can collect raw pixel data observed from dynamic environments and make decisions by itself based on the data. The decision making agent uses convolutional neural networks to classify situations it confronts, and the data observed from the environment undergoes preprocessing before being used. This research represents how the convolutional neural networks and the decision making agent are configured, analyzes learning performance through a value-based algorithm and a policy-based algorithm : a Deep Q-Networks and a Policy Gradient, sets forth their differences and demonstrates how the convolutional neural networks affect entire learning performance when using pixel data. This research is expected to contribute to the improvement of artificial intelligence systems which can efficiently find optimal solutions by using features extracted from raw pixel data.

• Korean Journal of Remote Sensing
• /
• v.23 no.3
• /
• pp.153-160
• /
• 2007

Forest cover density studies using high resolution satellite data and object oriented classification are limited in India. This article focuses on the potential use of QuickBird satellite data and object oriented classification in forest density mapping. In this study, the high-resolution satellite data was classified based on NDVI/pixel based and object oriented classification methods and results were compared. The QuickBird satellite data was found to be suitable in forest density mapping. Object oriented classification was superior than the NDVI/pixel based classification. The Object oriented classification method classified all the density classes of forest (dense, open, degraded and bare soil) with higher producer and user accuracies and with more kappa statistics value compared to pixel based method. The overall classification accuracy and Kappa statistics values of the object oriented classification were 83.33% and 0.77 respectively, which were higher than the pixel based classification (68%, 0.56 respectively). According to the Z statistics, the results of these two classifications were significantly different at 95% confidence level.

• Journal of the Korean Society of Visualization
• /
• v.21 no.1
• /
• pp.18-25
• /
• 2023

In these days, high dimensional data prediction technology based on neural network shows compelling results in many different kind of field including engineering. Especially, a lot of variants of convolution neural network are widely utilized to develop pixel level prediction model for high dimensional data such as picture, or physical field value from the sensors. In this study, velocity vector field of ideal flow on ship surface is estimated on pixel level by Unet. First, potential flow analysis was conducted for the set of hull form data which are generated by hull form transformation method. Thereafter, four different neural network with a U-shape structure were conFig.d to train velocity vectors at the node position of pre-processed hull form data. As a result, for the test hull forms, it was confirmed that the network with short skip-connection gives the most accurate prediction results of streamlines and velocity magnitude. And the results also have a good agreement with potential flow analysis results. However, in some cases which don't have nothing in common with training data in terms of speed or shape, the network has relatively high error at the region of large curvature.

• Proceedings of the KSRS Conference
• /
• v.2
• /
• pp.1011-1014
• /
• 2006

Coarse resolution (9 - 50 km pixels) Sea Surface Temperature satellite data are frequently considered adequate for open ocean research. However, coastal regions, including coral reef, estuarine and mesoscale upwelling regions require high-resolution (1-km pixel) SST data. The AVHRR SST data often suffer from navigation errors of several kilometres and still require manual navigation adjustments. The second serious problem is faulty and ineffective cloud-detection algorithms used operationally; many of these are based on radiance thresholds and moving window tests. With these methods, increasing sensitivity leads to masking of valid pixels. These errors lead to significant cold pixel biases and hamper image compositing, anomaly detection, and time-series analysis. Here, after manual navigation of over 40,000 AVHRR images, we implemented a new cloud filter that differs from other published methods. The filter first compares a pixel value with a climatological value built from the historical database, and then tests it against a time-based median value derived for that pixel from all satellite passes collected within ${\pm}3$ days. If the difference is larger than a predefined threshold, the pixel is flagged as cloud. We tested the method and compared to in situ SST from several shallow water buoys in the Florida Keys. Cloud statistics from all satellite sensors (AVHRR, MODIS) shows that a climatology filter with a $4^{\circ}C$ threshold and a median filter threshold of $2^{\circ}C$ are effective and accurate to filter clouds without masking good data. RMS difference between concurrent in situ and satellite SST data for the shallow waters (< 10 m bottom depth) is < $1^{\circ}C$, with only a small bias. The filter has been applied to the entire series of high-resolution SST data since1993 (including MODIS SST data since 2003), and a climatology is constructed to serve as the baseline to detect anomaly events.

• Journal of Broadcast Engineering
• /
• v.22 no.3
• /
• pp.350-365
• /
• 2017

Steganography is a technique for secret data communication, which is not perceived by third person between a receiver and a transmitter. It has been developed for thousands of years for the transmission of military, diplomatic or business information. The development of digital media and communication has led to the development of steganography techniques in modern times. Technic of image steganography include the LSB, which fixes the number of embedded bits into a pixel, and PVD, which exploits the difference value in the neighboring pixel pairs. In the case of PVD image steganography, a large amount of information is embedded fluidly by difference value in neighboring pixel pairs and the designed range table. However, since the secret information in order is embedded, if an error of the number of embedded bits occurs in a certain pixel pair, all subsequent information will be destroyed. In this paper, we proposes the method, which improve the vulnerability of PVD property about external attack or various noise and extract secret information. Experimental process is comparison analysis about stego-image, which embedded various noise. PVD shows that it is not possible to preserve secret information at all about noise, but it was possible to robustly extract secret information for partial noise of stego-image in case of the proposed PVD image steganography with locally-fixed number of embedding bits.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.11 no.7
• /
• pp.723-728
• /
• 2016

This study is analyzed the optical property of red tide pixel by using Landsat-7 ETM+, Landsat-8 OLI and COMS/GOCI image. In order to sample red tide pixel, Landsat-7, 8 true color image were used and obtained coordinate of red tide pixel in the true color image. Normalized water leaving radiance(nLw) and absorption coefficient were obtained from GOCI image in the same coordinate of the true color image. When red tide was not occurred the main absorption range was 412nm and 660nm but when red tide occurred it was 660nm and absorption coefficient in 412nm are drastically reduced. It made no difference of nLw spectrum between red tide pixel and non red tide pixel in nLw, but the absolute value of nLw was low than non red tide pixel, especially 660nm and 680nm wavelength sharply decrease.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.10
• /
• pp.1310-1316
• /
• 2021

Recently, with the development of artificial intelligence and IoT technology, the importance of video processing such as object tracking, medical imaging, and object recognition is increasing. In particular, the noise reduction technology used in the preprocessing process demands the ability to effectively remove noise and maintain detailed features as the importance of system images increases. In this paper, we provide a modified weight filter based on pixel matching in an AWGN environment. The proposed algorithm uses a pixel matching method to maintain high-frequency components in which the pixel value of the image changes significantly, detects areas with highly relevant patterns in the peripheral area, and matches pixels required for output calculation. Classify the values. The final output is obtained by calculating the weight according to the similarity and spatial distance between the matching pixels with the center pixel in order to consider the edge component in the filtering process.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.40 no.12
• /
• pp.2530-2538
• /
• 2015

The pixel value of depth image is depth value so that different objects which are placed on nearby position have similar pixel value. Moreover, the pixels of depth image have distinct pixel values compared to adjacent pixels while those of color image has very similar values. Accordingly distorted depth image of multiview video plus depth (MVD) needs proper error concealment methods considering the characteristics of depth image when transmission errors are happened. In this paper, classifying regions of depth image to consider edge directions and then applying adaptive error concealment methods to each region are proposed. Recovered depth images utilize with multiview video data to synthesize intermediate-view point video. The synthesized view is evaluated by objective quality metrics to demonstrate proposed method performance.

• Journal of the Institute of Electronics Engineers of Korea TE
• /
• v.37 no.3
• /
• pp.21-29
• /
• 2000

In this paper, we propose a way to detect a moving object efficiently and to track it using the active blocks. Instead of all Pixels in 8*8 Pixel value, any special pixel is extracted and we detect a moving object by comparison and by analysis the difference image information from darkness value of the same area. In the acquisition of image data by software processing, we reduce the number of data which obtain by convert high resolution image to low resolution image, and we can track a moving object in real time. So it can track a moving object in simple system without all the pixel value of the image data or additional VxD(Virtual x Driver). This system can be useful to track of a moving object in fixed block on PC(Personal Computer) and low custom.