Koshy Varghese

Use of Salient Features for the Design of a Multistage Framework to Extract Roads From High-Resolution Multispectral Satellite Images

The process of road extraction from high-resolution satellite images is complex, and most researchers have shown results on a few selected set of images. Based on the satellite data acquisition sensor and geolocation of the region, the type of processing varies and users tune several heuristic parameters to achieve a reasonable degree of accuracy. We exploit two salient features of roads, namely, distinct spectral contrast and locally linear trajectory, to design a multistage framework to extract roads from high-resolution multispectral satellite images. We trained four Probabilistic Support Vector Machines separately using four different categories of training samples extracted from urban/suburban areas. Dominant Singular Measure is used to detect locally linear edge segments as potential trajectories for roads. This complimentary information is integrated using an optimization framework to obtain potential targets for roads. This provides decent results in situations only when the roads have few obstacles (trees, large vehicles, and tall buildings). Linking of disjoint segments uses the local gradient functions at the adjacent pair of road endings. Region part segmentation uses curvature information to remove stray nonroad structures. Medial-Axis-Transform-based hypothesis verification eliminates connected nonroad structures to improve the accuracy in road detection. Results are evaluated with a large set of multispectral remotely sensed images and are compared against a few state-of-the-art methods to validate the superior performance of our proposed method.

Accelerometer-Based Activity Recognition in Construction

Recognizing the activities of workers helps to measure and control safety, productivity, and quality in construction sites. Automated activity recognition can enhance the efficiency of the measurement system. The present study investigates accelerometer-based activity classification for automating the work-sampling process. A methodology is developed for evaluating classifiers for recognizing activities based on the features generated from accelerometer data segments. An experimental study is carried out in instructed and uninstructed modes for classifying masonry activities by using accelerometers attached to the waist of the mason. Three types of classifiers were evaluated, and multilayer perceptron, a neural network classifier, gave the best results. A 50% overlap for data segments enhanced classifier performance. The study showed that the utilization of best features instead of all features did not affect the classification accuracy significantly but reduced the run time considerably. An accuracy of 8...

AUTOMATED PATH PLANNING FOR MOBILE CRANE LIFTS

Planning the lift path for a mobile crane involves the generation and selection of a suitable lift path through three-dimensional space while considering the degrees of freedom of the crane, its lifting capacity and potential site obstacles. This paper presents work done toward applying configuration space (C-space) and search concepts to develop a tool to identify lift paths that satisfies the planning requirements. An interference detection technique is used to generate the C-space and two levels of heuristic search are performed within the C-space. The first search is a heuristic depth search to determine the obstacle-free lift paths. The second search performs a more detailed optimization of the path within a constrained search space. The tool developed here can be used within the AutoCAD environment and is based on program modules developed using AutoLisp and external programs. Test findings indicate that the approach is capable of generating good paths in complex situations within reasonable time. Directions for further research are indicated.

A Heavy Lift Planning System for Crane Lifts

This paper presents the work done toward developing a computerized heavy lift planning system (HELPS) for planning crane lifts. Initially, a survey of the industry was carried out to define the heavy lift planning process and identify the developments necessary to improve the process. This survey identified eight tasks in the lift planning process. Based on these tasks, a logicalframework representing the planning process was developed. The scope of the current work was limited to developing a tool for planning three of these tasks. A visualization environment-Walkthru-was selected to implement the heavy lift planning system. Although Walkthru provided many of the features required for the system, critical functions had to be added. The developmental work for this study focused on (1) developing a shell that could control the visualization environment and related files to provide seamless access to the library of cranes and (2) providing features to perform critical lift planning functions. The resulting prototype system was tested on sample lifts, and all the functions worked as designed.

Selection of Accelerometer Location on Bricklayers Using Decision Trees

:  Preliminary investigations on accelerometer-based activity recognition in construction have shown that it has good potential to be utilized for recognizing categories of work in a construction trade. Selecting the accelerometer locations is an important consideration in activity recognition studies, but currently it is decided primarily on the basis of comfort requirements. This article proposes a methodology for selecting the location of accelerometers using video annotations and decision trees. A video annotation tool is used to track the movement of body segments and decision tree algorithm helps to prioritize the relevant body segments for classifying activities. A two-phase experimental study was conducted to assess the methodology. In the first phase, video annotation studies were carried out on four bricklayers, and based on decision tree analysis three locations: right lower arm, left lower arm, and waist were selected. In the second phase, an activity recognition study was conducted on another group of bricklayers with accelerometers attached at the selected locations. The results of study show that the location of accelerometer has a significant influence on accuracy and the proposed methodology is effective in selecting accelerometer locations. In the current study only bricklaying activity was considered, however, the methodology is generic and has the potential to be applied to objectively evaluate accelerator placement location for a wide range of structured activities.

AUTOMATED PROJECT SCHEDULE AND INVENTORY MONITORING USING RFID

Automated project monitoring system has the potential to be effective in controlling project performance through continuous monitoring and control of the project in real time using Automated Data Collection tools. This paper presents a system that is designed to investigate the use of RFID for automated schedule and inventory monitoring in real time. The system considers the use of precast concrete or prefabricated steel structural elements like columns, beams, floor slabs and wall panels as building components. The building components are identified in real time through automated data collection using RFID technology. This RFID technology with other application software is integrated to identify the installed component and update the project schedule in MSProject and present the as-built progress status. The 3D model of the building which is also updated is visualized in AutoCAD. The construction industry response towards the use of the developed system is also reported.

Formulating a coastal zone health metric for landuse impact management in urban coastal zones

The need for ICZM arises often due to inadequate or inappropriate landuse planning practices and policies, especially in urban coastal zones which are more complex due to the larger number of components, their critical dimensions, attributes and interactions. A survey of literature shows that there is no holistic metric for assessing the impacts of landuse planning on the health of a coastal zone. Thus there is a need to define such a metric. The proposed metric, CHI (Coastal zone Health Indicator), developed on the basis of coastal system sustainability, attempts to gauge the health status of any coastal zone. It is formulated and modeled through an expert survey and pertains to the characteristic components of coastal zones, their critical dimensions, and relevant attributes. The proposed metric is applied to two urban coastal zones and validated. It can be used for more coast friendly and sustainable landuse planning/masterplan preparation and thereby for the better management of landuse impacts on coastal zones.

INVESTIGATION OF AN AHP BASED MULTI CRITERIA WEIGHTING SCHEME FOR GIS ROUTING OF CROSS COUNTRY PIPELINE PROJECTS

Pipelines are the most efficient, cost effective and environmentally friendly means of fluid transport [1]. When selecting the route of a cross country pipeline, the goal is to route it so that it has the greatest utility to the public while minimizing negative impacts to people and the natural environment. In order to accomplish this goal of addressing several factors, a large amount of location based information needs to be analyzed. Hence the manual routing processes are very tedious in nature. The advent of new techniques like that of Geographic Information System (GIS) has made the routing process of pipelines more systematic and effective. This work identifies the various factors which need to be considered while routing a pipeline in the Indian scenario. A GIS to represent the various factors and automatically route the pipeline has been developed. A main drawback in the system is the lack of a structured methodology to derive the relative preferences of the different factors affecting the route. To address this issue, an AHP based methodology was developed. The GIS and the AHP methodology have been tested on case studies in the Indian scenario.

A Structured Approach to Form Dependency Structure Matrix for Construction Projects

The sequence of design activities is vital, particularly in interdisciplinary projects. Sequencing of activities is influenced by the information dependency among the activities. The dependency relationships among the design activities in a project are of three types - independent, dependent and interdependent or loops. Interdependent activities (or loops) require assumed information to start and progress. If these assumptions are erroneous, it can lead to rework of a single or a group of dependent activities. One of the challenges in planning the sequencing of design activities is to decide on appropriate information or assumptions which can be made to break the interdependent activities (or loops). Conventional tools such as Critical Path Method (CPM) and Program Evaluation and Review Technique (PERT) are not suitable for representing information flow because interdependent activities or loops cannot be modeled. Dependency Structure Matrix (DSM) has been identified as a potential tool to model feedback loops/ cycles. One of the main difficulties in implementing DSM is in comprehensively identifying the activities and information dependencies which are required to formulate the DSM. Based on the experience in attempting to develop a DSM for an interdisciplinary project, the authors have proposed a new approach in identifying all the possible activities and their information dependencies. A prototype software also had been developed in Excel VBA for implementing the methodology. The proposed concepts are illustrated through application to Induced Draft Cooling Tower (IDCT) projects. This illustration is multidisciplinary in nature consisting of mechanical, civil, electrical and instrumentation. The proposed concept has proven fruitful and hence this approach is anticipated beneficial for construction applications.

A Hierarchical Multi-classifier Framework for Landform Segmentation Using Multi-spectral Satellite Images - A Case Study over the Indian Subcontinent

There is an increasing need for automatically segmenting the regions of different landforms from a multispectral satellite image. The problem of Landform classification using data only from a 3-band optical sensor (IRS-series), in the absence of DEM (Digital Elevation Model) data, is complex due to overlapping and confusing spectral reflectance from several different landform classes. We propose a hierarchical method for landform classification for identifying a wide variety of landforms occurring over parts of the Indian subcontinent. At the first stage, the image is classified into one of three broad categories: Desertic, Coastal or Fluvial, using decision fusion of three SVMs (Support Vector Machine). In the second stage, the image is then segmented into different regions of landforms, specifically belonging to the class (category) identified at stage 1. To show the improvement in accuracy of our classification method, the results are compared with two other methods of classification.