K. Sunil Kumar

Joint segmentation and image interpretation

Abstract The problem of image interpretation is formulated in the framework of modular integration and multiresolution. The formulation essentially involves the concept of reductionism and multiresolution, where the image interpretation task is broken down into simpler subtasks of segmentation and interpretation. Moreover, instead of solving the vision task at the finest resolution Ω, we solve the synergetically coupled vision subtasks at coarser resolutions Ω −ξ for Ω ⩾ξ>0 and use the results obtained at resolution ( Ω −ξ) to solve the vision task at resolution ( Ω −ξ+1) . We present a solution to the joint segmentation and interpretation problem in the proposed framework. For the interpretation part we exploit the Markov random field (MRF) based image interpretation scheme developed by Modestino and Zhang [IEEE Trans Pattern Anal. Mach. Intell. pp. 606–615 (1992)]. Experimental results on both indoor and outdoor images are presented to validate the proposed framework.

New algorithms for 3D surface description from binocular stereo using integration

Abstract In this paper, we formulate and develop an approach which integrates different modules (feature extractor, matching and interpolation) involved in stereo. We study the integration process at the finest resolution when (i) the precomputed edge map is the only line field driving the model, (ii) the line fields are computed interactively by the feature extracting module of the model, and (iii) when both the interactive line field computation module and the precomputed line field modules are present. This integration process being computationally intensive, we develop a multiresolution stereo integration approach. The energy function for each module at different resolutions is constructed and minimized in an integrated manner yielding a dense disparity map. A new energy function for the matching module is proposed. Experimental results are presented to illustrate our approach.

An integrated method using AHP, DEA and GP for evaluating supply sources

The paper relates to a study carried out to develop an integrated framework capable of serving as a decision support system to deal with the problem of outsourcing. Initially, a survey was carried out among the Indian industries which revealed that most of the firms followed specific procedures for outsourcing and followed simple qualitative procedures for source selection. However, there was no evidence that they used any quantitative-based decision making tools for justifying outsourcing, selection of vendors, etc. The framework developed is capable of taking into account both qualitative and quantitative factors in making outsourcing decisions. The framework uses multi stage procedure involving analytic hierarchy process (AHP), data envelopment analysis (DEA) and preemptive goal programming (PGP). At the outset, the framework uses an AHP-based methodology to arrive at a decision whether to outsource or not and if outsourced, whom to outsource. This is followed by a DEA-based procedure, wherein the relative efficiencies of the priority sources, identified by AHP are found. Finally, an AHP-PGP formulation identifies the best source for the given targets of the outsourcer. The details of the framework and how was it used with specific cases are presented in this paper.

A multiresolution approach to color image restoration and parameter estimation using homotopy continuation method

We address the problem of color image restoration. We model the image as a Markov random field (MRF) and propose a restoration algorithm in a multiresolution framework. The incorporation of multiresolution technique significantly reduces the computational complexity of the restoration algorithm. The energy function at each resolution being non-convex, is minimized using the simulated annealing algorithm. The parameters which describe the MRF model at each resolution are computed a priori using the homotopy continuation method. Simulation results are presented to validate the proposed scheme.

Locating human faces in a cluttered scene

In this paper, we present two new schemes for finding human faces in a photograph. The first scheme adopts a distribution-based model approach to face-finding. Distributions of the face and the face-like manifolds are approximated using higher order statistics (HOS) by deriving a series expansion of the density function in terms of the multivariate Gaussian and the Hermite polynomials in an attempt to get a better approximation to the unknown original density function. An HOS-based data clustering algorithm is then proposed to facilitate the decision process. The second scheme adopts a hidden Markov model (HMM) based approach to the face-finding problem. This is an unsupervised scheme in which face-to-nonface and nonface-to-face transitions are learned by using an HMM. The HMM learning algorithm estimates the HMM parameters corresponding to a given photograph and the faces are located by examining the optimal state sequence of the HMM. We present experimental results on the performance of both schemes. A training data base of face images was constructed in the laboratory. The performances of both the proposed schemes are found to be quite good when measured with respect to several standard test face images.

Prospecting rural supply chain management using a structured approach

The Indian retail, manufacturing and infrastructure sectors are poised for rapid growth and are facing new challenges in logistics and supply chain issues related to rural areas. Various reports suggest that with over 70% of the economy centred around the rural parts of the country, these markets offer tremendous scope and many firms are looking at appropriate strategies to address this issue. However, many practitioners and researchers seem to believe that the issues related to marketing and supply chain are radically different from the traditional approaches hitherto used. It was envisaged to verify how realistic these opinions are. The approach used was to develop structural relationships involving three well known frameworks such as the supply chain operations reference (SCOR) elements, marketing processes (MP) and the marketing mix variables (MMV) in the context of rural supply chain so as to evolve strategies for understanding and managing rural supply chain better. The study revealed that there is significantly varying associations among different groups of marketing processes with the elements of SCOR and this is also true with different groups of marketing variables. The findings of the study are reported in this paper.

The perspectives and practices of business process outsourcing in Indian industries: a survey and analysis

Outsourcing has become a common terminology among the industries, which has provided a thrust in achieving the organisational goals. Organisations are narrowing the range of processes and functions they perform internally, and todays business environment relies substantially on third party business and service providers. This trend is very much on the increase in India. A survey of the Indian industries was carried out to get an insight into the level of maturity of the companies as far as outsourcing is concerned. The survey covered issues related to various facets of outsourcing including the functions outsourced, reasons for outsourcing, benefits obtained, problems faced and criteria for vendor selection. The response obtained from more than 100 participants drawn from various sectors, including manufacturing, information technology and service, was subjected to preliminary analysis and multivariate statistical test such as factor analysis. The results of these analyses provide significant insight into the perspectives and practices of business process outsourcing (BPO) in Indian industries, especially in terms of the relationships that seem to exist among different functions outsourced, criteria for preference among the vendors, levels of advantages offered by BPO and the degree of problems faced by outsourcing organisations. The results also provide a good base to develop an effective decision support system. The details of the survey, the analysis and the key learning points are presented in this paper.

Profiling Indian rural markets for supply chain management

The Indian rural market is set to become a billion dollar market in the years to come. With over 70% of the economy centred on the rural parts of the country, the marketers are now seeking ways to address this issue, as immense potential is latent in the rural parts of the country. In order to understand the rural supply chain (RSC) better, an attempt was made to identify the factors that are pertinent in the Indian context. After identifying these factors, data was collected pertaining to various dimensions that would provide an indication to these factors. The data was collected for rural regions of the major states in India. The data for the rural districts of a particular state was also collected as a sample so as to conduct a focused study. Appropriate tools were used to group the states of the country and the districts of that state. The results thus obtained would help marketers to evolve strategies looking at the communalities involved among the rural regions, the details of which are presented in this paper.

Mapping structural relationships among the critical factors of rural supply chains

The Indian retail, manufacturing and infrastructure sectors are poised for rapid growth but they are faced with new challenges when it comes to the untapped rural areas. Rural India is experiencing significant changes due to the pace of economic development in India. This includes changing consumption patterns, increasing exposure to different lifestyles and products, and increasing purchasing power. In order to understand the rural supply chain (RSC) better, an attempt was made to identify the factors that are pertinent in the Indian context. The factors have a profound impact in designing a suitable supply chain model for the rural markets. After identifying these factors, interpretive structural modelling (ISM) and fuzzy analysis was used to establish structural relationships among these factors to categorise them into sensitive and influential sets. The results obtained revealed a number of interesting facts about RSC, the details of which are presented in this paper.

A modular integration and multiresolution framework for image interpretation

In this paper, we give a generalized formulation for a vision problem in the framework of modular integration and multiresolution. The developed framework is used to solve the high-level vision problem of scene interpretation. The formulation essentially involves the concept of reductionism and multiresolution, where the given vision taskν is broken down into simpler subtasksν1,ν2, …,νm. Moreover, instead of solving the vision taskνΩ =ν at the finest resolution Θ, we solve the synergetically coupled vision subtasks at coarser resolutions (Ω −N) for Θ ≥N>0 and use the results obtained at resolution (Θ −N) to solveνΩ −N+1, the vision task at resolution (Θ −N+1). Image interpretation is a two-phased analysis problem where some salient features or objects in an image are identified by segmenting the image and the objects in the segmented image are interpreted based on their spatial relationships. We present a solution to the joint segmentation and interpretation problem in the proposed generalized framework. For the interpretation part we exploit the Markov Random Field (MRF) based image interpretation scheme developed by Modestino and Zhang. Experimental results on both indoor and outdoor images are presented to validate the proposed framework.