Traditional Segmentation Approaches Research Articles

Online Semantic Object Segmentation for Vision Robot Collected Video

Video segmentation is the foundation of many senior computer vision applications. For robotic vision applications, video object segmentation is facing more difficulties. Since the videos are usually collected from those complicated scenes and the camera perspective may be static for some time, object saliency and motion clues which are widely used in traditional video segmentation approaches could be unreliable. Besides, its online processing ability is of crucial importance. In this paper, we propose an online video object segmentation method, which is organized as a segmentation-by-detection framework. It first effectively segments out meaningful objects in short video frame batches with the help of object detectors, and then the segmentation of different batches can be associated with each other via bi-directional notebook based connecting strategy. Experiments conducted on public available datasets verify the above-mentioned property and show good performance of the proposed method.

The application of convolution neural network based cell segmentation during cryopreservation

For most of the cells, water permeability and plasma membrane properties play a vital role in the optimal protocol for successful cryopreservation. Measuring the water permeability of cells during subzero temperature is essential. So far, there is no perfect segmentation technique to be used for the image processing task on subzero temperature accurately. The ice formation and variable background during freezing posed a significant challenge for most of the conventional segmentation algorithms. Thus, a robust and accurate segmentation approach that can accurately extract cells from extracellular ice that surrounding the cell boundary is needed. Therefore, we propose a convolutional neural network (CNN) architecture similar to U-Net but differs from those conventionally used in computer vision to extract all the cell boundaries as they shrank in the engulfing ice. The images used was obtained from the cryo-stage microscope, and the data was validated using the Hausdorff distance, means ± standard deviation for different methods of segmentation result using the CNN model. The experimental results prove that the typical CNN model extracts cell borders contour from the background in its subzero state more coherent and effective as compared to other traditional segmentation approaches.

Stacked fully convolutional networks with multi-channel learning: application to medical image segmentation

The automated segmentation of regions of interest (ROIs) in medical imaging is the fundamental requirement for the derivation of high-level semantics for image analysis in clinical decision support systems. Traditional segmentation approaches such as region-based depend heavily upon hand-crafted features and a priori knowledge of the user. As such, these methods are difficult to adopt within a clinical environment. Recently, methods based on fully convolutional networks (FCN) have achieved great success in the segmentation of general images. FCNs leverage a large labeled dataset to hierarchically learn the features that best correspond to the shallow appearance as well as the deep semantics of the images. However, when applied to medical images, FCNs usually produce coarse ROI detection and poor boundary definitions primarily due to the limited number of labeled training data and limited constraints of label agreement among neighboring similar pixels. In this paper, we propose a new stacked FCN architecture with multi-channel learning (SFCN-ML). We embed the FCN in a stacked architecture to learn the foreground ROI features and background non-ROI features separately and then integrate these different channels to produce the final segmentation result. In contrast to traditional FCN methods, our SFCN-ML architecture enables the visual attributes and semantics derived from both the fore- and background channels to be iteratively learned and inferred. We conducted extensive experiments on three public datasets with a variety of visual challenges. Our results show that our SFCN-ML is more effective and robust than a routine FCN and its variants, and other state-of-the-art methods.

Clustering Algorithm Based on Spatial Shadowed Fuzzy C-means and I-Ching Operators

In this paper, the authors are devoted to design a new segmentation approach based on I-Ching operators in the framework of shadowed fuzzy C-means clustering. The I-Ching operators are innovative operators, which are evolved from ancient Chinese I-Ching philosophy. I-Ching operators include three kinds of operators, intrication operator, turnover operator, and mutual operator. These new operators are very flexible and efficient in evolution procedure. In this paper, the new operators are specifically designed to search for the optimal cluster centers of shadowed fuzzy C-means. Considering the local spatial information in image segmentation procedure, a new segmentation algorithm called I-Ching spatial shadowed fuzzy C-means (ICSSFCM) is proposed. Traditional segmentation approaches based on fuzzy C-means, shadowed fuzzy C-means, and spatial shadowed fuzzy C-means are compared with the proposed method. The experimental results show that the proposed ICSSFCM is very efficient approach not only in tackling the overlapping segments but also in suppressing the noise in images.

Optic disc segmentation by balloon snake with texture from color fundus image.

A well-established method for diagnosis of glaucoma is the examination of the optic nerve head based on fundus image as glaucomatous patients tend to have larger cup-to-disc ratios. The difficulty of optic segmentation is due to the fuzzy boundaries and peripapillary atrophy (PPA). In this paper a novel method for optic nerve head segmentation is proposed. It uses template matching to find the region of interest (ROI). The method of vessel erasing in the ROI is based on PDE inpainting which will make the boundary smoother. A novel optic disc segmentation approach using image texture is explored in this paper. A cluster method based on image texture is employed before the optic disc segmentation step to remove the edge noise such as cup boundary and vessels. We replace image force in the snake with image texture and the initial contour of the balloon snake is inside the optic disc to avoid the PPA. The experimental results show the superior performance of the proposed method when compared to some traditional segmentation approaches. An average segmentation dice coefficient of 94% has been obtained.

Segmentation of desert sand dunes

Dune fields occupy considerable land areas not only on earth but also on other planetary surfaces like Venus, Mars and Saturn’s moon Titan. Their characteristic orientations pose a challenge for studying them using remote sensing imagery. In this letter, a novel methodology to segment sand dunes as observed on satellite imagery is proposed. The Radon spectrum for an image is obtained by using various projections sampled at finer angular directions. Statistical features such as mode, median, mean, standard deviation and entropy have been defined to characterize the textural information captured by Radon spectrum. Landsat and LISS-III data were used to test the utility of the methodology developed. Comparisons with traditional segmentation approaches have demonstrated the effectiveness of the proposed methodology. It is seen that mode, entropy and standard deviation features of Radon spectrum are more suitable for segmenting dune regions.

High-Resolution Remote Multi-Spectral Sensing Images Based on Texture Features

Aiming at the difficulties in the segmentation for high-resolution remote multispectral sensing images, this paper proposed a segmentation approach for remote sensing images based on texture features. The algorithm implemented precipitation watershed transform respectively on the texture images obtained by the different characteristics of GLCM, and then superimposed the two segmentation results, finally completing the image segmentation by using a novel regional consolidation method that combined the texture features. The experiments were implemented on the high-resolution ALOS and SPOT 5 remote sensing images respectively. Compared with the traditional watershed segmentation approach based on gradient information, the experimental results showed that the proposed algorithm can accurately locate the edges of objects, effectively overcome the phenomenon of over-segmentation and under-segmentation, with a higher segmentation accuracy and stability.

Name segmentation using hidden Markov models and its application in record linkage.

This study aimed to evaluate the use of hidden Markov models (HMM) for the segmentation of person names and its influence on record linkage. A HMM was applied to the segmentation of patient's and mother's names in the databases of the Mortality Information System (SIM), Information Subsystem for High Complexity Procedures (APAC), and Hospital Information System (AIH). A sample of 200 patients from each database was segmented via HMM, and the results were compared to those from segmentation by the authors. The APAC-SIM and APAC-AIH databases were linked using three different segmentation strategies, one of which used HMM. Conformity of segmentation via HMM varied from 90.5% to 92.5%. The different segmentation strategies yielded similar results in the record linkage process. This study suggests that segmentation of Brazilian names via HMM is no more effective than traditional segmentation approaches in the linkage process.

Variety-seeking: Using the CHAID segmentation approach in analyzing the international traveler market

This study investigates the use of variety-seeking in the segmentation of international travelers. Segmentation studies mainly utilize traditional segmentation criteria, such as nationality, country of origin, other trip characteristics of international travelers, and consumer behaviors. Variety-seeking has proven to be a relevant factor for a wider understanding of the behavior of international travelers. Hence, segmentation that is based on variety-seeking offers an alternative to the traditional segmentation approaches. The use of chi-squared automatic interaction detection (CHAID) identified several segments based on variety-seeking behavior. Findings are discussed within the framework of variety-seeking literature. Managerial implications are provided based on this new way of segmenting the market of international travelers.

Recognising customers' patterns of bus service patronage using a loyalty framework

With the aim of identifying typical characteristics of travellers, traditional segmentation approaches were based on socio-demographic variables. However, the approaches could not reveal the factors motivating individual behaviour. This result led to an emerging interest in psychological research models that are adhered to the decision-making process. Among various related theories, the concept of loyalty was attractive because the major purpose of establishing a loyalty concept is to recognise a customer's pattern towards a given service. However, there were few efforts aimed at determining market segments based on a loyalty framework. In addition, there was no consensus achieved on theoretical loyalty typology due to different empirical findings in different market contexts. This study aims to be the first loyalty-based attempt to provide an operational method of segmenting bus service market. Seeking practical implementation, another focus of this study is to determine typical characteristics of the market segments. Analyses that included cluster techniques were conducted on questionnaire data collected from 333 respondents in Hidaka city, Japan. A cross-classification between relative attitude and service patronage was successfully established, dividing the market into four segments. Segments of loyalty and no-loyalty were observed to be dominant over the remaining market. In contrast, the spurious loyalty segment was small and insignificant. An expansion of the latent loyalty segment was also observed when moving from the intention phase to the actual behaviour phase. Notably, not only demographic factors but also social awareness variables including environmental concern and elderly support were observed to be significant in distinguishing customer segments from one another.

Hidden Markov random field model based brain MR image segmentation using clonal selection algorithm and Markov chain Monte Carlo method

The hidden Markov random field (HMRF) model has been widely used in image segmentation, as it provides a spatially constrained clustering scheme on two sets of random variables. However, in many HMRF-based segmentation approaches, both the latent class labels and statistical parameters have been estimated by deterministic techniques, which usually lead to local convergence and less accurate segmentation. In this paper, we incorporate the immune inspired clonal selection algorithm (CSA) and Markov chain Monte Carlo (MCMC) method into HMRF model estimation, and thus propose the HMRF–CSA algorithm for brain MR image segmentation. Our algorithm employs a three-step iterative process that consists of MCMC-based class labels estimation, bias field correction and CSA-based statistical parameter estimation. Since both the MCMC and CSA are global optimization techniques, the proposed algorithm has the potential to overcome the drawback of traditional HMRF-based segmentation approaches. We compared our algorithm to the state-of-the-art GA–EM algorithm, deformable cosegmentation algorithm, the segmentation routines in the widely-used statistical parametric mapping (SPM) software package and the FMRIB software library (FSL) on both simulated and clinical brain MR images. Our results show that the proposed HMRF–CSA algorithm is robust to image artifacts and can differentiate major brain structures more accurately than other three algorithms.

Emerging segmentation practices in the age of the social customer

This paper reports on an exploratory study of segmentation practices of organisations with a social media presence. It investigates whether traditional segmentation approaches are still relevant in this new socio-technical environment and identifies emerging practices. The study found that social media are particularly promising in terms of targeting influencers, enabling the cost-effective delivery of personalised messages and engaging with numerous customer segments in a differentiated way. However, some problems previously identified in the segmentation literature still occur in the social media environment, such as the technical challenge of integrating databases, the preference for pragmatic rather than complex solutions and the lack of relevant analytical skills. Overall, a gap has emerged between marketing theory and practice. While segmentation is far from obsolete in the age of the social customer, it needs to adapt to reflect the characteristics of the new media.

A (mis)guided adventure tourism experience: An autoethnographic analysis of mountaineering in Bolivia

Due to the fast growing nature of the adventure tourism industry and the commodification of adventure activities therein, improved understanding of adventure tourism experiences and mountaineer adventure tourists in particular is needed. In an effort to move beyond traditional market segmentation approaches, this study analysed autoethnographical data from an adventure tourism mountaineering experience in Bolivia. This autoethnographic method facilitated a deeper understanding of mountaineering adventure tourism experiences and allowed for a multifaceted view of risk perceptions that has often been neglected in the literature. Data were analysed with a robust psychological framework (i.e. reversal theory) that was used to explain: (a) paradoxical desires for risk and safety in adventure tourism and (b) emotional and motivational fluctuations experienced by mountaineer adventure tourists. The importance of creating a ‘protective frame’ to ensure enjoyable experiences was identified, along with key factors that influenced this frame (e.g. guide behaviour, equipment, safety management procedures, other tourists, environmental conditions). Implications for adventure tourism practitioners are discussed, along with theoretical analyses. The utility of autoethnographic research in adventure settings, particularly in conjunction with established psychological theory, is highlighted and suggested as a fruitful avenue through which to enhance the adventure tourism discourse.

Computing with Biologically Inspired Neural Oscillators: Application to Colour Image Segmentation

This paper investigates the computing capabilities and potential applications of neural oscillators, a biologically inspired neural model, to grey scale and colour image segmentation, an important task in image understanding and object recognition. A proposed neural system that exploits the synergy between neural oscillators and Kohonen self-organising maps (SOMs) is presented. It consists of a two-dimensional grid of neural oscillators which are locally connected through excitatory connections and globally connected to a common inhibitor. Each neuron is mapped to a pixel of the input image and existing objects, represented by homogenous areas, are temporally segmented through synchronisation of the activity of neural oscillators that are mapped to pixels of the same object. Self-organising maps form the basis of a colour reduction system whose output is fed to a 2D grid of neural oscillators for temporal correlation-based object segmentation. Both chromatic and local spatial features are used. The system is simulated in Matlab and its demonstration on real world colour images shows promising results and the emergence of a new bioinspired approach for colour image segmentation. The paper concludes with a discussion of the performance of the proposed system and its comparison with traditional image segmentation approaches.

Hierarchical Segmentation for Hardware Function Evaluation

This paper presents a method for evaluating functions based on piecewise polynomial approximations (splines) with a hierarchical segmentation scheme targeting hardware implementation. The methodology provides significant reduction in table size compared to traditional uniform segmentation approaches. The use of hierarchies involving uniform splines and splines with size varying by powers of two is particularly well suited for the coverage of nonlinear regions. The segmentation step is automated and supports user-supplied precision requirements and approximation method. Bit-widths of the coefficients and arithmetic operators are optimized to minimize circuit area and enable a guarantee of 1 unit in the last place (ulp) accuracy at the output. A coefficient transformation technique is also described, which significantly reduces the dynamic ranges of the fixed-point polynomial coefficients. The hierarchical segmentation method is illustrated using a set of functions including -(x/2) log <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> x, cos <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-1</sup> (x), radic(-ln( <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">x</i> )) , a high-degree rational function, ln(1+x), and 1/(1+x). Various degree-1 and degree-2 approximation results for precisions between 8 to 24 bits are given. Hardware realizations are demonstrated on a Xilinx Virtex-4 field-programmable gate array (FPGA).

Methodological and Strategy Development Implications of Decision Segmentation

The goal of customer segmentation, the cornerstone of strategy development, is to identify homogeneous groups of customers that will respond in a consistent way to changes in the marketing mix. Interpretation of traditional quantitative segmentation approaches requires an inferential leap as to the underlying decision processes of each segment. Means-end research methodologies address this problem by providing a framework to understand customer decision making that can be directly translated into the specification of positioning strategy that is more personally relevant to a given target consumer group. The quantitative marketing research orientation to means-end research is contrasted to a more qualitative, consumer-decision research perspective. A new methodological procedure that addresses the shortcomings in previous analysis methods to produce decision segments is presented.

Segmenting Customers from Population to Individuals: Does 1-to-1 Keep Your Customers Forever?

There have been various claims made in the marketing community about the benefits of 1-to-1 marketing versus traditional customer segmentation approaches and how much they can improve understanding of customer behavior. However, few rigorous studies exist that systematically compare these approaches. In this paper, we conducted such a study and compared the predictive performance of aggregate, segmentation, and 1-to-1 marketing approaches across a broad range of experimental settings, such as multiple segmentation levels, multiple real-world marketing data sets, multiple dependent variables, different types of classifiers, different segmentation techniques, and different predictive measures. Our experiments show that both 1-to-1 and segmentation approaches significantly outperform aggregate modeling. Reaffirming anecdotal evidence of the benefits of 1-to-1 marketing, our experiments show that the 1-to-1 approach also dominates the segmentation approach for the frequently transacting customers. However, our experiments also show that segmentation models taken at the best granularity levels dominate 1-to-1 models when modeling customers with little transactional data using effective clustering methods. In addition, the peak performance of segmentation models are reached at the finest granularity levels, skewed towards the 1-to-1 case. This finding adds support for the microsegmentation approach and suggests that 1-to-1 marketing may not always be the best solution

The advantages of preference-based segmentation: An investigation of online grocery retailing

Most companies aim to identify different groups of attractive customers in order to offer them appropriate products and/or services. To do this, companies need market segmentation. There is, however, a problem with the standard methods employed in market segmentation. The static inductive approach to market segmentation commonly employed by companies does little to identify buying intentions across demographic variables. A more dynamic and deductive approach to developing market segmentation is through analysis of consumer preference structures — an approach which is glaringly absent from most marketing texts, not least because of the difficulty of developing practical approaches which can generate effective marketing strategies. The purpose of this paper is to highlight and demonstrate a preference-based approach to market segmentation, using shoppers' experience of online grocery retail brands in the UK. The paper first demonstrates how using choice-based conjoint analysis could help achieve this objective more effectively than other more traditional conjoint methods. While conjoint analysis is not new, the ability to segment based on markedly different preference structures is a recent development and comprises a powerful, but underutilised segmentation approach. A web-based methodology is then applied to build up a picture of consumers' conscious and unconscious prioritisation of a large number of choice criteria. The paper calculates consumer utility values for both offline and online shoppers in the UK and develops a preference-based segmentation approach, which is compared with a traditional demographic segmentation approach using the same data. From this analysis, the advantages of a preference-based approach to segmentation are extracted. The paper closes with recommendations for market research practitioners.

Situational segmentation in the international marketplace: the Japanese snack market

The study demonstrates the viability of situational segmentation in a market outside the USA. A number of situational segmentation studies in the USA have examined the snacking market. This study examines situational segmentation opportunities in the context of the Japanese snacking market. The study attempts to delineate a situationally‐defined market structure for a broadly defined array of snack products. This is done by characterizing 18 snacks in terms of pertinent situational factors via dummy variable regression analysis; and grouping the snacks in terms of the similarity of their situational characterizations via cluster analysis. The study reveals four multi‐product snack segments, including solitary snacking cluster, socializing ensemble cluster, high gravity socialization cluster, and morning home snack. The results show that situational segmentation is as effective in complementing more traditional segmentation approaches in Japan as it is in the USA.

Consumers' decision-making style as a basis for market segmentation

Decision-making styles are important to marketing because they determine consumer behaviour, are relatively stable over time and thus are relevant for market segmentation. The need to test the generalisability of Sproles and Kendall's consumer styles inventory (CSI) in different countries and an attempt to extend the original work led the authors to test the structure of decision-making styles of German shoppers and its use in segmenting consumers. From a sample of 455 German consumers, a seven-dimensional structure of decision-making styles was found using principal components analysis and confirmatory factor analysis. Cluster analysis identified six meaningful and distinct decision-making groups. These decision-making groups can be viewed as basic segments that can be used in conjunction with traditional market segmentation approaches. Implications for marketing research and practitioners are discussed.