Object-Oriented Approach of Information Extraction from High Resolution Satellite Imagery

Abstract. Monitoring and analyzing the (decreasing) trends in lake freezing provides important information for climate research. Multi-temporal satellite images are a natural data source to survey ice on lakes. In this paper, we describe a method for lake ice monitoring, which uses low spatial resolution (250 m–1000 m) satellite images to determine whether a lake is frozen or not. We report results on four selected lakes in Switzerland: Sihl, Sils, Silvaplana and St. Moritz. These lakes have different properties regarding area, altitude, surrounding topography and freezing frequency, describing cases of medium to high difficulty. Digitized Open Street Map (OSM) lake outlines are back-projected on to the image space after generalization. As a pre-processing step, the absolute geolocation error of the lake outlines is corrected by matching the projected outlines to the images. We define the lake ice detection as a two-class (frozen, non-frozen) semantic segmentation problem. Several spectral channels of the multi-spectral satellite data are used, both reflective and emissive (thermal). Only the cloud-free (clean) pixels which lie completely inside the lake are analyzed. The most useful channels to solve the problem are selected with xgboost and visual analysis of histograms of reference data, while the classification is done with non-linear support vector machine (SVM). We show experimentally that this straight-forward approach works well with both MODIS and VIIRS satellite imagery. Moreover, we show that the algorithm produces consistent results when tested on data from multiple winters.

The document summarizes the steps, procedures, information infrastructure used for collecting built up information from satellite imagery. The document is intended as a reference for visual interpreters. The document also summarizes the lessons learned by photo-interpreters for imagery covering Europe.

Spatially distributed actual evapotranspiration (ETa) val- ues were estimated based on satellite data and the Surface Energy Balance System (SEBS) approach for the Karkheh River Basin, Iran. Nineteen cloud free MODIS (Moderate Resolution Imaging Spec- troradiometer) images, representing a complete cropping year from November 2002 to October 2003 were acquired and processed. Esti- mated ETa values were verified using sub-catchment scale water bal - ance analysis. The results revealed that during the study period ETa was estimated at 16680x10 6 m 3 and that the water balance closure terms at sub-basin scale ranged from 0.6% to 7.2% of precipitation. This implies that water balance is sufficiently understood. Estimated outflow from the basin was 7.8% of precipitation and indicates that water is a very scarce resource in the Karkheh. Rain fed areas con- sume about 3720 x10 6 m 3 /year and are mainly located in the sub- catchments of the upper Karkheh while irrigated areas consume 2680x 10 6 m 3 /year and are mainly located in the lower areas in the basin. Total water consumption by forest is about 2070×10 6 m 3 /year, mainly in the middle parts of the basin. The range lands are scattered mainly all over the Upper Karkheh and together with areas in the Lower Karkheh consume about 3360×10 6 m 3 /year. ETa from other land uses is 4110×10 6 m 3 /year, of which ETa from open water surfac- es is the main contributor. The Karkheh Dam evaporates 80×10 6 m 3 / year while wetlands located in the lower area of the basin evaporate 660×10 6 m 3 /year. Satellite data along with the SEBS algorithm and geo-statistical techniques are effective to estimate spatial patterns of water consumption and availability. These facilitate the introduction of diverse management interventions to different areas in the basin based on the real ground situation.

In addition to meteorological inputs, land use and emission factor database are also required for simulation of the Biogenic Emission Inventory System (BEIS). The resolution of land use database in the Taiwan BEIS (TBEIS) can be up to 1 km^2 and this database consists of vegetation species and their spatial distribution in Taiwan with a total land area of approx 36000 km^2. The previous Taiwanese land use database was established in 1995, but since then Taiwan has gone through cultural, ecological and seismic changes for more than a decade. As a result, the area and distribution of current land use types needed revision. To complete this revision accurately, recent satellite remote sensing data were used in this research. In the updated land use database, the new forested area had increased by 2400 km^2, grass land had increased by 1790 km^2, paddy fields had increased by 2,100 km^2 and dry agriculture land had decreased by 1830 km^2. There was a net increase of 4460 km^2 in vegetation area, which equates to about 16% of the total vegetation land area of Taiwan. Simulation of the TBEIS-2 rendered to an increase of 28,000 ton/yr in total BVOCs (6.6%), in which other VOCs (OVOC) increased by 17,000 ton/yr (12.1%) and isoprene increased by 10,000 ton/yr (7.9%). The biogenic emissions obtained by the use of TBEIS-2 with the new land use database are utilized to simulate air quality by the Taiwan Air Quality Monitoring (TAQM). Results indicated that concentrations of ozone and other photochemical pollutants raised in some areas of Taiwan. In the simulation conducted in May 2003, the greatest ozone concentration difference occurred in southern Taiwan and central Taiwan with an increase of 3 ~ 7 ppb in these districts.

The availability of productive land becomes one of the main problems in the development of food security and independence of Indonesian farmers. After the regional autonomy era begins, the food business is the responsibility of local government, both at provincial and district / city, the implementation and operational excellence tailored to the potential for their respective regions. Act No. 7 of 1996 on Food, in particular article 45, 46, 47, and 48 explain that achieve food security is the responsibility of the government and the community. In order to support making the right decisions to maintain local food security, support is needed more accurate data, especially in raw paddy fields as the main source of producing rice. Extensive data collection of raw paddy fields along with other supporting data, basically to increase the availability of extensive data and information materials on various levels of administrative fields that have a high accuracy. Satellite imagery interpretation that can produce high accuracy is high-resolution satellite imagery such as Ikonos or image Quickbird image. Based on the results of this mapping is concluded that the total land area of raw rice is obtained from measurements in four districts (Kec. Barabai, Kec. Labuan Amas Selatan, Kec. Labuan Amas Utara and Kec. Haruyan) in Middle River Upper District is 25,277.35 ha consisting of 9,311.17 ha of rainfed lowland, 1,876.00 and 14,090.18 ha irrigated lowland.

شناخت آسیب پذیری جمعیت و گروه های انسانی به عنوان اقدامی ابزاری در راستای پیشگیری از زایش فاجعه مبتنی بر زیستن در نواحی پر خطر مطرح است. تحلیل آسیب پذیری کانون‌های جمعیتی واقع در عرصه های خطر اقدامی علمی برای فراهم سازی زیر ساخت اطلاعاتی مناسب از درجه و میزان خطرپذیری از رخداد مخاطره است. بر اساس چنین دیدگاهی مسئله تحلیل آسیب پذیری ناشی از مخاطرات ژئومورفولوژیک فرونشست در جنوب شهر تهران مورد توجه قرار گرفته است. بدین منظور در ابتدا با استفاده از تکنیک تداخل سنجی راداری نرخ و دامنه فرونشست در محدوده مورد مطالعه استخراج گردید. سپس برای تولید نقشه فضایی عرصه های پر مخاطره نرخ و دامنه فرونشست به عنوان فاکتور عامل در زایش مخاطره به تحلیل گر سیستم در محیط GIS وارد گردید. نقشه خروجی معرف سلول‌های واجد ارزش متفاوت از فاکتور زایش مخاطره است. استخراج عرصه متعلق پهنه در معرض مخاطره نشان داد که از مجموع مساحت محدوده 7/12 درصد در پهنه های بسیار پرخطر، 7/14 در عرصه پرخطر،5/15 در عرصه نسبتاً پرخطر بیش از نیمی از محدوده (57درصد) در عرصه نسبتاً کم خطر قرار گرفته‌اند. برای ارزیابی میزان خطرپذیری، الگوی توزیع فضایی تأسیسات و ساخت‌وساز های موجود با نقشه عرصه های در معرض خطر هم نهاد گردید. مبتنی بر این اطلاعات و با استفاده از پیمایش میدانی اقدام به بررسی آسیب پذیری جمعیت ،مساکن و تأسیسات موجود در محدوده مورد مطالعه صورت پذیرفت. یافته های تحقیق موید آسیب پذیر بودن اکثر سکونت گاه‌ها و جمعیت ساکن درآنهاو تأسیسات موجود در محدوده می‌باشد. همچنین بر اساس پایش های میدانی صورت گرفته مشخص گردید که اکثر ساخت و سازهای موجود در منطقه غیر اصولی بوده که این امر باعث فراهم آوری زیرساخت آسیب پذیری برای مستحدثات در شرایط خطرپذیر است.

This paper proposes a fuzzy model of emotion and behaviours selection for an autonomous mobile robot. Conventional behaviour-based robots can only exhibit a single behaviour or action after sensing an external event. In order to make a robot to adapt to its environment and to exhibit more complex behaviours, we try to model animal emotions and behaviours by a set of Fril's support logic fuzzy rules. This set of fuzzy rules will allow the robot to control and to select different behaviours depending on the environment and its experience. An summary of the calculus of support logic is given in this paper.

For more than two decades, Mars has been of great interest for scientific exploration, using orbiting spacecrafts or landers. Orbiters like “Mars Express” cover large areas and give images of the surface with a resolution limited to several meters. Rovers missions, like “Pathfinder” or “MER”, offer the ability to maneuver to scientifically interesting sites, perform analysis of soil and rock composition, but are limited in range to the immediate surroundings of the landing site. There is a strategic gap for systems that would combine high-resolution imagery and extensive coverage. This gap can be addressed by micro-airplanes that will give more accurate images than satellites, thanks to their proximity to the ground (about 0.5 to 2 km) and cover much larger areas than rovers without being limited by terrain roughness. This paper presents a feasibility study and the first design of a very lightweight solar powered airplane for Mars exploration. Called Sky-Sailor, it was investigated in an ESA definition study within the framework of the Startiger technology program.

Currently in IRAN, some manual methods are used for measuring spray droplet size which is very laborious and time consuming. In this paper, a software is presented based on image processing techniques, which is suitable for small object detection purpose that can be used in blob analysis of sprayer calibration and other similar applications. The algorithms have been implemented and adapted in a program called Scientific Image Blob Analysis (SIBA). The program can read the scanned image of water sensitive cards, obtained from nozzle's spray. Some threshold techniques are also used for noise removal and converting the original image to a binary one (e.g. red for blob and white for background). The blob features calculated in the program are area, actual diameter, Number Median Diameter (NMD), Volume Median Diameter (VMD), blob density (blob/unit area), and percentage of blob's area coverage, blob count and finally the uniformity of spray. The algorithms were evaluated analytically and experimentally, comparing with the results of a leaf area meter and a planimeter. Although the results showed more accurate and faster performance for this software, a direct correlation was also found between their accuracy and the image's resolution. In other words, the higher the resolution of the image, the more accurate of the results particularly for the smaller blobs. This was mainly due to the stepwise inherent of the pixels showing a blob's boundary. However, by choosing an appropriate resolution the error of measurement may reduce to %1. The program can also easily be used for any area detection and calculation of the images such as leaf, fruits etc. It is a good package especially for the researchers in the developing countries with a limit access to similar resources. Moreover there might be some uncertainty about the accuracy of such commercial products. Key words: Blob Analysis, Image processing, water sensitive paper, droplet size

The purpose of this research was to examine high school adminastrators' conflict resolution strategies and tactics from a social constructivist perspective. Participants were 709 students, 75 teachers, and 21 school administrators from 7 high schools located in low socioeconomic areas in Izmir. Data were collected through adminestrating two different questionnaire and an interview protocol. Participants were selected through stratified, simple rondom, and convenionce sampling techniques. Results showed that there were differences among students, teachers and school administrators' views regarding school managers' conflict resoluiton strategies and tactics. Unlike the students, teachers stated that school managers used more constructive conflict resolution strategies and tactics rather than destructive ones. Whereas 62 % of teachers claimed that school managers used constructive conflict resolution strategies, only 27 % of the students shared this claim. Similarly more female students stated that school managers used constructive conflict resolution strategies and tactics than did the male students. School administrators' perceptions about themselves were radically diffirent from the students' perceptions about school administrators. Although teachers and especially female students stated that school administrators used constructive conflict resolution strategies and tactics, these strategies and tactics were limited to the use of giving advice and talking students. When whole results of the research were examined based on social constructivist theory, it can be said that school managers' conflict resolution strategies and tactics can not lead to students' construction of knowledge and skills concerning conflict resolution strategies.

Reproduction couleur par trames irrégulières et semi-régulières

Fuzzy rules automatically derived from a set of training examples quite often produce better classification results than fuzzy rules translated from medical knowledge. This study aims to investigate the difference in domain representation between a knowledge-based and a data-driven fuzzy system applied to an electrocardiography classification problem. For a three-class electrocardiographic arrhythmia classification task a set of fifteen fuzzy rules is derived from medical expertise on the basis of twelve electrocardiographic measures. A second set of fuzzy rules is automatically constructed on thirty-nine MIT-BIH database's records. The performances of the two classifiers on thirteen different records are comparable and up to a certain extent complementary. The two fuzzy models are then analyzed, by using the concept of information gain to estimate the impact of each ECG measure on each fuzzy decision process. Both systems rely on the beat prematurity degree and the QRS complex width and neglect the P wave existence and the ST segment features. The PR interval is not well characterized across the fuzzy medical rules while it plays an important role in the data-driven fuzzy system. The T wave area shows a higher information gain in the knowledge based decision process, and is not very much exploited by the data-driven system. The main difference between a human designed and a data driven ECG arrhythmia classifier is found about the PR interval and the T wave.

Southern elephant seals (Mirounga leonina) play a pivotal role in the Southern Ocean as wide-ranging marine predators and major prey consumers within Southern Ocean marine ecosystems. Due to their circumpolar distribution and the remoteness of their habitat, large uncertainties remain about their total population sizes. This is especially true for elephant seal populations in the French Southern Territories in the southern Indian Ocean (i.e. Crozet and Kerguelen Archipelagos) as many breeding sites are inaccessible for ground censuses. Here, we present a simple and efficient approach for estimating the total elephant seal populations of the Kerguelen and Crozet Archipelagos by using very high-resolution satellite imagery (<1m resolution). Twenty-eight satellite images taken during the breeding season to count female elephant seals in inaccessible areas were used and complemented the traditional annual ground counts in accessible areas. For Kerguelen Island sectors likely to host colonies and where no satellite images were available for the breeding season, a statistical predictive model was built to estimate the most likely number of breeding females to be present on a given beach according to its physiographic characteristics. Our results show the reliability of using very high-resolution satellite images, a relatively low-cost platform, to count pinniped populations and provide the first estimation of the total southern elephant seal population for both the Kerguelen 347,995 (s e = 4,950) and Crozet 13,065 (s e = 169) Archipelagos. The combined total represents over 35% of the global elephant seal population with the Kerguelen stock being numerically equivalent to the South Georgia stock. In addition, we re-examined the population trends since the last mid-century for Kerguelen and over the last five decades for Crozet. The demographic trends of the southern Indian Ocean populations show marked growth over the last decade (5.1% and 1.6% annual growth rate for Crozet and Kerguelen respectively), particularly on Crozet where the elephant seal population has more than tripled.

Rule extraction from data is one of the key technologies for solving the bottlenecks in artificial intelligence. Artificial neural networks are well suited for representing any knowledge in given data. Extraction of logical/fuzzy rules from the trained artificial neural network is of great importance to researchers in the fields of artificial intelligence and soft computing. Fuzzy rule sets are capable of approximating any nonlinear mapping relationships. Extraction of rules from data has been discussed in terms of fuzzy modeling, fuzzy clustering, and classification with fuzzy rule sets. This special issue entitled"Rule Extraction from Data" is aimed at providing the readers with good insights into the advanced studies in the field of rule extraction from data using neural networks/fuzzy rule sets. I invited seven research papers best suited for the theme of this special issue. All the papers were reviewed rigorously by two reviewers each. The first paper proposes an interesting rule extraction method from data using neural networks. Ishikawa presents a combination of learning with an immediate critic and a structural learning with forgetting. This method is capable of generating skeletal networks for logical rule extraction from data with correct and wrong answers. The proposed method is applied to rule extraction from lense data. The second paper presents a new methodology for logical rule extraction based on transformation of MLP (multilayered perceptron) to a logical network. Duck et al. applied their C-MLP2LN to the Iris benchmark classification problem as well as real-world medical data with very good results. In the third paper, Geczy and Usui propose fuzzy rule extraction from trained artificial neural networks. The proposed algorithm is implied from their theoretical study, not from heuristics. Their study enables to initially consider derivation of crisp rules from trained artificial neural network, and in case of conflict, application of fuzzy rules. The proposed algorithm is experimentally demonstrated with the Iris benchmark classification problem. The fourth paper presents a new framework for fuzzy modeling using genetic algorithm. The authors have broken new ground of fuzzy rule extraction from neural networks. For the fuzzy modeling, they have proposed a particular type of neural networks containing nodes representing membership functions. In this fourth paper, the authors discuss input variable selection for the fuzzy modeling under multiple criteria with different importance. A target system with a strong nonlinearity is used for demonstrating the proposed method. Kasabov, et al. present, in the fifth paper, a method for extraction of fuzzy rules that have different level of abstraction depending on several modifiable thresholds. Explanation quality becomes better with higher threshold values. They apply the proposed method to the Iris benchmark classification problem and to a real world problem. J. Yen and W. Gillespie address interpretability issue of Takagi-Sugeno-Kang model, one of the most popular fuzzy mdoels, in the fifth paper. They propose a new approach of fuzzy modeling that ensures not only a high approximation of the input-output relationship in the data, but also good insights about the local behavior of the model. The proposed method is applied to fuzzy modeling of sinc function and Mackey-Glass chaotic time series data. The last paper discusses fuzzy rule extraction from numerical data for high-dimensional classification problems. H.Ishibuchi, et al. have been pioneering methods for classification of data using fuzzy rules and genetic algorithm. In this last paper, they introduced a new criterion, simplicity of each rule, together with the conventional ones, compactness of rule base and classification ability, for high-dimensional problem. The Iris data is used for demonstrating their new classification method. They applied it also to wine data and credit data. I hope that the readers will be encouraged to explore the frontier to establish a new paradigm in the field of knowledge representation and rule extraction.

We have been using fuzzy data mining techniques to extract patterns that represent behavior for intrusion detection. We describe a variety of modifications that we have made to the data mining algorithms in order to improve accuracy and efficiency. We use sets of fuzzy association rules that are mined from network audit data as models of behavior. To detect anomalous behavior, we generate fuzzy association rules from new audit data and compute the similarity with sets mined from normal data. If the similarity values are below a threshold value, an alarm is issued. We describe an algorithm for computing fuzzy association rules based on Borgelt's (2001) prefix trees, modifications to the computation of support and confidence of fuzzy rules, a new method for computing the similarity of two fuzzy rule sets, and feature selection and optimization with genetic algorithms. Experimental results demonstrate that we can achieve better running time and accuracy with these modifications.