Parallelepiped Classification Research Articles

Comprehensive Land Use and Land Cover Classification in Chhatrapati Sambhajinagar City, Maharashtra, India, Using Satellite Imagery and Machine Learning Techniques

Abstract: In This research presents an extensive analysis of land use and land cover (LULC) classification in Chhatrapati Sambhajinagar City, a rapidly expanding urban centerin Maharashtra, India, highly susceptible to climate change. Using remote sensing and GIS techniques, the research evaluates the impact of LULC on ecosystem service values (ESV) over a nine- year period (2013-2021). Satellite-based data were utilized to quantify LULC changes, encompassing Green Area, Buildup Area, Road Area, WaterBodies, and Barren Land. The research employs five distinct classification methods and investigates the effectiveness of pre-processing techniques in improving classification accuracy during selected years (2013, 2016, 2019, and 2021). The performance of Mahalanobis Distance, Maximum Likelihood, Parallelepiped, and Support Vector Machine methods is assessed under both preprocessed and non-preprocessed conditions, using overall accuracy and Kappa accuracy as evaluation metrics. The results reveal consistent enhancements in overall accuracy, with Mahalanobis Distance, Maximum Likelihood, and Support Vector Machine methods demonstrating percentage improvements ranging from 0.43% to 5.08% after preprocessing. Particularly noteworthy are the substantial percentage increases in Parallelepiped Classification, ranging from 23.74% to 53.34%, highlighting the transformative impact of preprocessing on its accuracy. These findings emphasize the necessity for customized preprocessing strategies to refine the precision of LULC classification models, offering valuable contributions to the realms of remote sensing and geospatial analysis. Moreover, the insights gleaned from this research provide crucial guidance for urban growth management and climate resilience strategies in rapidly developing urban areas, making it a valuable resource for sustainable urban planning

Modelling the relationship between climatic variables and land use/land cover classes in Yewa South Local Government Area of Ogun State, Nigeria

Climate change and land use/land cover (LULC) change are among the global environmental issues that have recently dominated international debates in order to monitor earth resources. The burning of fossils fuels, clearing of land and burning of bush has given rise to the highest levels of greenhouse pollution in the atmosphere. It is on this note that this study was carried out to determine the effect of climate change on land use/land cover change of Yewa South Local Government Area of Ogun State in Nigeria. The supervised parallelepiped classification scheme was used to classify land use land cover into five classes namely agriculture land, built-up, vegetation, wetland and water body. A multivariate linear regression model was developed and program in MATLAB to model the relationship between climatic variables namely, rainfall (mm), average temperature (°C), evaporation (mm), relative humidity (%), average soil temperature (°C) and some particular land use/land cover variables. After statistical analysis between climatic variables and land use/land cover classification, results indicated a significant relationship between climatic variables and land use/land cover change. The statistical analysis between the variables shows that the climatic parameters has an accuracy of 84.12% prediction model for land use/land cover classification.

Computer-aided detection of bone metastasis in bone scintigraphy images using parallelepiped classification method.

Accurate diagnosis of metastatic tissue on bone scintigraphy images is of paramount importance in making treatment decisions. Although several automated systems have developed, more and better interpretation methods are still being sought. In the present study, a new modality for bone metastasis detection from bone scintigraphy images using parallelepiped classification (PC) as method for mapping the radionuclide distribution is presented. Bone scintigraphy images from 12 patients with bone metastases were analyzed using the parallelepiped classifier that generated color maps of scintigraphic images. Seven classes of radionuclide accumulation have been identified and fed into machine learning software. The accuracy of the proposed method was evaluated by statistical measurements in a confusion matrix. Overall accuracy, producer's and user's accuracies and κ coefficient were computed from each confusion matrix associated with the individual case. The results revealed that the method is sufficiently precise to differentiate the metastatic bone from normal tissue (overall classification accuracy = 87.58 ± 2.25% and κ coefficient = 0.8367 ± 0.0252). The maps are easier to read (due to better contrast) and can detect even slightest differences in accumulation levels among pixels. In conclusion, these preliminary data suggest that bone scintigraphy combined with PC method could play an important role in the detection of bone metastasis, allowing for an easier but correct interpretation of the images, with effects on the diagnosis accuracy and decision making on the treatment to be applied.

Spatial-Temporal Change Assessment of Manchar and Keenjhar Lake Using RS and GIS

To maintain sustainability of life on earth, water is an indispensable nutrient and truly a vital resource for aquatic ecosystem. It also plays an important role in the socio-economic development and is one of the key factors for driving the economy of any region. Rapid increase in population and change in climatic patterns has vigorously affected the water resources. The remote sensing is an efficient tool for monitoring of natural resources. This research is focused on spatio-temporal change analysis of Manchar Lake and Keenjhar Lake, Sindh, Pakistan using Landsat multi-sensor data MSS, TM, ETM+ & OLI to analyze land cover from 1972 to 2017 and calculating water indices using Principal Component Analysis (PCA) and supervised parallelepiped classification achieving an overall accuracy 98.86% and 0.1 of standard deviation value. The results revealed that the significant changes in land cover due to unusual climatic patterns, significant fall in precipitation rate and multiple anthropogenic activities. These changes had adverse impact on eco-system of lakes and as well as on public health

COMPARISON OF SUPERVISED CLASSIFICATION TECHNIQUES WITH ALOS PALSAR SENSOR FOR ROORKEE REGION OF UTTARAKHAND, INDIA

Abstract. The Advanced Land Observing Satellite (ALOS) is developed by the Japanese Aerospace Exploration Agency (JAXA) which was launched in the year 2006 for the Earth observation and exploration purpose. The ALOS was carrying PRISM, AVNIR-2 and PALSAR sensors for this purpose. PALSAR is L-Band synthetic aperture radar (SAR). The PALSAR sensor is designed in a way that it can work in all weather conditions with a resolution of 10 meters. In this research work we have made an investigation on the accuracy obtained from the various supervised classification techniques. We have compared the accuracy obtained by classifying the ALOS PALSAR data of the Roorkee region of Uttarakhand, India. The training ROI’S (Region of Interest) are created manually with the assistance of ArcGIS Earth and for the testing purpose, we have used the Global positioning system (GPS) coordinates of the region. Supervised classification techniques included in this comparison are Parallelepiped classification (PC), Minimum distance classification (MDC), Mahalanobis distance classification (MaDC), Maximum likelihood classification (MLC), Spectral angle mapper (SAM), Spectral information divergence (SID) and Support vector machine (SVM). Later, through the post classification confusion matrix accuracy assessment test is performed and the corresponding value of the kappa coefficient is obtained. In the result, we have concluded MDC as best in term of overall accuracy with 82.3634% and MLC with a kappa value of 0.7591. Finally, a peculiar relationship is developed in between classification accuracy and kappa coefficient.

Performance of Spectral Angle Mapper and Parallelepiped Classifiers in Agriculture Hyperspectral Image

Hyperspectral Imaging (HSI) is used to provide a wealth of information which can be used to address a variety of problems in different applications. The main requirement in all applications is the classification of HSI data. In this paper, supervised HSI classification algorithms are used to extract agriculture areas that specialize in wheat growing and get a classified image. In particular, Parallelepiped and Spectral Angel Mapper (SAM) algorithms are used. They are implemented by a software tool used to analyse and process geospatial images that is an Environment of Visualizing Images (ENVI). They are applied on Al-Kharj, Saudi Arabia as the study area. The overall accuracy after applying the algorithms on the image of the study area for SAM classification was 66.67%, and 33.33% for Parallelepiped classification. Therefore, SAM algorithm has provided a better a study area image classification.

고해상도 수치항공정사영상기반 하천토지피복지도 제작을 위한 분류기법 연구

The information on the land cover along stream corridor is important for stream restoration and maintenance activities. This study aims to review the different classification methods for mapping the status of stream corridors in Seom River using airborne RGB and CIR digital ortho imagery with a ground pixel resolution of 0.2m. The maximum likelihood classification, minimum distance classification, parallelepiped classification, mahalanobis distance classification algorithms were performed with regard to the improvement methods, the skewed data for training classifiers and filtering technique. From these results follows that, in aerial image classification, Maximum likelihood classification gave results the highest classification accuracy and the CIR image showed comparatively high precision.

A decision-tree classification for low-lying complex land cover types within the zone of discontinuous permafrost

This study presents a decision-tree (DT) approach to classifying heterogeneous land cover types within a northern watershed located in the zone of discontinuous permafrost using airborne LiDAR and high resolution spectral datasets. Results are compared with a more typically applied supervised classification. Increasing errors in discharge resulting from an inaccurate classification are quantified using a distributed hydrological model.The hierarchical classification was accurate between 88% and 97% of the validation sub-area, whereas the parallelepiped classification was accurate between 38% and 74% of the same area (despite overall accuracy of ~91%, kappa=0.91). Topographical derivatives were best able to explain variations in land cover types (82% to 96%), whilst spectral and vegetation structural derivatives were less accurate. When compared with field measurements, the hierarchical classification of plateau edges (adjacent to a fen) was within 2m of measured, 60% of the time, whilst this occurred only 40% of the time when using a spectral classification. When examining the impacts of land cover classification accuracy on modelled discharge, we find that the length of the Hydrological Response Unit defined by the classification (and subject to varying levels of errors) was linearly related to discharge (m3) such that an increase in permafrost plateau area would increase discharge by 26% of the total. The methodology presented in this paper clarifies previous classification and modelling studies using Landsat and IKONOS data for the same basin. This study greatly improves upon past classifications in the same area, furthers our understanding of the distribution of connected bogs and fens (as conveyors of water to the basin outlet) within the watershed, and current spatial extents of rapidly thawing permafrost plateaus, which are critical for better understanding the impacts of climate change on these northern environments.

An Entropy-Based Multispectral Image Classification Algorithm

Employing the entropy theory, this paper presents a new and robust multispectral image classification algorithm. The digital number (DN) in remotely sensed multispectral images is considered as a random variable when judging the allocation of unknown pixels into predefined training classes. If an unknown pixel shows a similar DN vector as the pixels in a training class, it will increase the global entropy defined as the sum of DN probabilities multiplied by the logarithm of DN probabilities for all pixels within the training class. The unknown pixel is to be assigned to the class for which the entropy of the training class is increased most due to the inclusion of the pixel. The proposed entropy-based classification (EC) is compared with the maximum likelihood classification (MLC), parallelepiped classification, minimum distance classification, Mahalanobis distance classification (MDC), iterative self-organizing data analysis technique (ISODATA) classification, and K-means classification. These classifiers were applied to a Landsat Enhanced Thematic Mapper Plus image covering Houston, Texas, USA, acquired on October 16, 1999. A reference land cover map from the National Land Cover Data 2001 of the same area was taken as a ground reference to assess the accuracy of classification results, suggesting that the EC showed comparable overall accuracy as MDC, and they both outperformed other classifiers. The results of MLC can be improved by substituting the multivariate lognormal or gamma distribution for the multivariate normal distribution involved in its assumption. The EC algorithm has the potential to produce reliable land cover maps regardless of the distribution of DN vectors and relevant parameters of probability density functions involved in other classifiers.

Detection of Hidden Bruise on Kiwi fruit Using Hyperspectral Imaging and Parallelepiped Classification

Abstract It is necessary to develop non-destructive detection techniques of kiwi fruit because machine injury could lower the quality of fruit and incur economic losses. Owing to the special physical properties of kiwi fruit peel, the bruises are not visible externally. Its could not be effectively inspected using conventional non-destructive detection technology. We proposed the hyperspectral imaging technique to inspect the hidden bruises on kiwi fruit in this work. The Vis/NIR (408-1117 nm) hyperspectral image data was collected. The top four component images were obtained from the data which ranged from 600 to 1000 nm using principal component analysis, and the bruise regions were extracted from the component images using parallelepiped classification. The experimental results show that the error of detecting hidden bruises on fruits with hyperspectral imaging was 14.5%.

Integrated use of remote sensing and geographic information systems in riparian vegetation delineation and mapping

This paper presents a pilot study on riparian vegetation delineation and mapping using remote sensing and geographic information systems (GIS) in the Hunter Region, Australia. The aim of the study was to develop appropriate and repeatable assessment and mapping techniques to quantify the extent of riparian vegetation within the region. Ortho‐rectified digital aerial photographs, SPOT‐4 multispectral (XS) and Landsat‐7 ETM+ images were tested to delineate the riparian vegetation and to develop a quantifiable and repeatable method of mapping the extent of that vegetation. Image processing techniques such as parallelepiped classification, tasselled cap transformation, and vegetation index clustering were used in an attempt to delineate riparian vegetation from remotely sensed images. Specific GIS analysis techniques were used for riparian zone buffering and segmentation, vegetation cover estimation, and mapping. Specific GIS scripts were developed for those processes so that they were automatic, fast, and repeatable. Various vegetation indices (VI) were evaluated and compared for their ability to discriminate riparian vegetation and its extent. The riparian vegetation was assessed and mapped at designed segmental interval (e.g. 1 km) with polygon and lineal representations. The classification accuracy was assessed against field observation and air photo interpretation (API). The overall accuracy of the photo‐based classification was about 81%, SPOT‐4 63%, and Landsat‐7 ETM+ 53%. Statistic analysis shows that there is little agreement between photo‐based classification and that from satellite imagery.

Use of IKONOS and Landsat for malaria control in the Republic of Korea

Malaria reemerged in the Republic of Korea (ROK) in 1993. While limited numbers of U.S. soldiers in high-risk areas use chloroquine/primaquine chemoprophylaxis to prevent malaria, control of mosquito larvae through larviciding also can be used to reduce the risk of malaria transmission. In order to estimate the cost of larviciding, accurate estimates of the spatial extent of mosquito larval habitats are necessary. The purpose of this study was to determine whether an accurate estimate of the area covered by mosquito larval habitats can be obtained using Landsat 7 Enhanced Thematic Mapper+ (ETM+) and/or IKONOS data for the Korean test site. To estimate the area covered by larval habitats near Camp Greaves [Paekyeon-Ri, near Tongil-Chon (village)] in the ROK, an IKONOS and a Landsat 7 ETM+ image were classified using a parallelepiped classification. In a comparison with rice paddy field sites, 24 (92%) of the sites were classified correctly on the IKONOS image and 17 (65%) were classified correctly on the Landsat image. Comparing the classifications on a pixel-by-pixel basis, the agreement between the two classifications was 79%. Part of the disagreement was due to the difference in resolution of the two images. In spite of local differences, the two classifications produced similar area estimates. Although either Landsat or IKONOS could be used in Korea for a reasonable estimate of habitat area, only IKONOS can resolve small irrigation ponds. While ponds represent a small portion of the total larval habitat, they are an important source for mosquito breeding during the late rice-growing season in the ROK since they contain higher larval densities. High-resolution imagery, such as IKONOS, would be necessary for planning and implementing treatment of these smaller habitats.

Regional Climatic Similarities in the Temperate Zones of North and South America

We performed an analysis of the climatic patterns of the temperate zones in North and South America using a global database of monthly precipitation and temperature. Three synthetic variables, identified by a principal component analysis (PCA) of the monthly data, were used: mean annual precipitation, mean annual temperature and the proportion of the precipitation falling during summer. We displayed the spatial gradient of the three variables by constructing a com- posite colour raster image. We used a parallelepiped classification algorithm to locate areas in both continents that are climatically similar to five North American Long Term Ecological Research sites and to two South American long- term ecological research sites. The same algorithm was used to identify areas in South America which are climatically similar to some of the major grassland and shrubland types of North America. There is substantial overlap between the climates of North and South America. Most of the climatic patterns found in South America are well represented in North America. How- ever, there are certain climates in North America that are not found in South America. An example is a climate with relatively low mean annual temperature and high summer precipitation. The climatic signatures of three North American LTER sites (Cedar Creek, CPER and Sevilleta) were not found in South America. The climatic signatures of two LTER sites (Konza and Jornada) had some representation in South America. Two South American research sites (Rio Mayo and Las Chilcas) were well represented climatically in North America. The climates of six out of seven selected North American grassland and shrubland types were represented in South America. The northern mixed prairie type was not represented climatically in South America. Our analysis sug- gests that comparisons of North and South America can provide a powerful test of climatic control over vegetation.

RGB-NDVI colour composites for visualizing forest change dynamics

The study presents a simple and logical technique to display and quantify forest change using three dates of satellite imagery. The normalized difference vegetation index (NDVI) was computed for each date of imagery to define high and low vegetation biomass. Color composites were generated by combining each date of NDVI with either the red, green, or blue (RGB) image planes in an image display monitor. Harvest and regeneration areas were quantified by applying a modified parallelepiped classification creating an RGB-NDVI image with 27 classes that were grouped into nine major forest change categories. Aerial photographs and stand history maps are compared with the forest changes indicated by the RGB-NDVI image. The utility of the RGB-NDVI technique for supporting forest inventories and updating forest resource information systems are presented and discussed.

Digital Land Cover Mapping Using Layered Classification Logic and Physical Composition Attributes

Imagery created by satellite multispectral scanning systems have generated widespread interest for application to land cover mapping. The energy sensed in each spectral band can be interpreted visually, or it can be used as a feature in a machine-assisted pattern recognition process. Simultaneous visual interpretation of numerous multispectral images often becomes impractical, especially as the area studied increases; hence, great emphasis has been placed on computer techniques for improving the repeatability, accuracy, and timeliness of land cover mapping. In this light, a parallelepiped layered classification incorporating physical composition attributes was developed and tested against the traditional Boolean logic parallelepiped decision rule. LANDSAT digital picture elements (pixels) of Goleta, California, were interactively filtered through a series of decision layers (solid, liquid, organic, inorganic, mixture solid/liquid, and mixture organic/inorganic) prior to being assigned to land use classes. Evaluation of the layered classification versus the traditional parallelepiped classification revealed a statistically significant difference between errors of commission and no significant difference between errors of omission. Errors of commission create serious problems in traditional parallelepiped classifications, consequently the hybrid technique exhibits potential for improved land cover mapping.