We propose a theoretically and practically improved density-based, hierarchical clustering method, providing a clustering hierarchy from which a simplified tree of significant clusters can be constructed. For obtaining a “flat” partition consisting of only the most significant clusters (possibly corresponding to different density thresholds), we propose a novel cluster stability measure, formalize the problem of maximizing the overall stability of selected clusters, and formulate an algorithm that computes an optimal solution to this problem. We demonstrate that our approach outperforms the current, state-of-the-art, density-based clustering methods on a wide variety of real world data.

Density-based clustering based on hierarchical density estimates

Postgraduate Students: Vladimír Arnot, Daniel Čapek, Rudolf Čejka, Dao Minh, Tomá Dulík, Martin Hrubý, Radek Kočí, Petr Kotásek, Marek Křejpský, Bohuslav Křena, Vladislav Kubíček, Vladimír Marek, Petr Matouek, Ale Mičín, Jiří Očenáek, Tomá Ondráček, Petr Peňás, Jaroslav Ráb, Richard Růička, Luká Sekanina, Ivan Schwarz, Azedien Sllame, Petr Smolík, Jiří Staroba, Josef Strnadel, Luká Szemla, Pavel Tinovský, Michal Tomů, Milan Urbáek, Michal Vojkůvka, Petr Vurm, Frantiek Zbořil

Department of Computer Science and Engineering

Color histograms lack spatial information and are sensitive to intensity variation, color distortion and cropping. As a result, images with similar histograms may have totally different semantics. The region-based approaches are introduced to overcome the above limitations, but due to the inaccurate segmentation, these systems may partition an object into several regions that may have confused users in selecting the proper regions. In this paper, we present a robust image retrieval based on color histogram of local feature regions (LFR). Firstly, the steady image feature points are extracted by using multi-scale Harris-Laplace detector. Then, the significant local feature regions are ascertained adaptively according to the feature scale theory. Finally, the color histogram of local feature regions is constructed, and the similarity between color images is computed by using the color histogram of LFRs. Experimental results show that the proposed color image retrieval is more accurate and efficient in retrieving the user-interested images. Especially, it is robust to some classic transformations (additive noise, affine transformation including translation, rotation and scale effects, partial visibility, etc.).

Robust image retrieval based on color histogram of local feature regions

Aiming at the problemsof too much iterative times in selecting initial centroids stochastically for K-Means algorithm,a method is proposed to optimize the initial centroids through cutting the set into k segmentations and select one point in each segmentation as initial centroids for iterative computing. A new valid function called clustering-index is defined as the sum of clustering-density and clustering-significance and can be used to search the optimization of k in the internal of [1,n(1/2) ]. The simulation experiment with IRIS data set shows that the proposed algorithm converges faster and the value k found is close to the actual value,which proves the validity of the algorithm.

An Improved K-Means Clustering Algorithm

Image semantic segmentation has always been a research hotspot in the field of robots. Its purpose is to assign different semantic category labels to objects by segmenting different objects. However, in practical applications, in addition to knowing the semantic category information of objects, robots also need to know the position information of objects to complete more complex visual tasks. Aiming at a complex indoor environment, this study designs an image semantic segmentation network framework of joint target detection. Using the parallel operation of adding semantic segmentation branches to the target detection network, it innovatively implements multi-vision task combining object classification, detection and semantic segmentation. By designing a new loss function, adjusting the training using the idea of transfer learning, and finally verifying it on the self-built indoor scene data set, the experiment proves that the method in this study is feasible and effective, and has good robustness.

https://ietresearch.onlinelibrary.wiley.com/doi/pdf/10.1049/iet-ipr.2020.0088

Jointly network image processing: multi-task image semantic segmentation of indoor scene based on CNN

https://cyberleninka.org/article/n/1019246.pdf

Texture classification and discrimination for region-based image retrieval

Semantic image segmentation is a fundamental yet challenging problem, which can be viewed as an extension of the conventional object detection with close relation to image segmentation and classification. It aims to partition images into non-overlapping regions that are assigned predefined semantic labels. Most of the existing approaches utilize and integrate low-level local features and high-level contextual cues, which are fed into an inference framework such as, the conditional random field (CRF). However, the lack of meaning in the primitives (i.e., pixels or superpixels) and the cues provides low discriminatory capabilities, since they are rarely object-consistent. Moreover, blind combinations of heterogeneous features and contextual cues exploitation through limited neighborhood relations in the CRFs tend to degrade the labeling performance. This paper proposes an ontology-based semantic image segmentation (OBSIS) approach that jointly models image segmentation and object detection. In particular, a Dirichlet process mixture model transforms the low-level visual space into an intermediate semantic space, which drastically reduces the feature dimensionality. These features are then individually weighed and independently learned within the context, using multiple CRFs. The segmentation of images into object parts is hence reduced to a classification task, where object inference is passed to an ontology model. This model resembles the way by which humans understand the images through the combination of different cues, context models, and rule-based learning of the ontologies. Experimental evaluations using the MSRC-21 and PASCAL VOC'2010 data sets show promising results.

/pdf/ontology-based-semantic-image-segmentation-using-mixture-5c0exayds3.pdf

Ontology-Based Semantic Image Segmentation Using Mixture Models and Multiple CRFs.

A novel technique for Content-Based Image Retrieval (CBIR) that employs both the color and spatial information of images is proposed. A maximum of three dominant color regions in an image together with its respective coordinates of the Minimum-Bounding Rectangle (MBR) are first extracted. Next, the Sub-Block technique is then used to determine the location of the dominant regions by comparing the coordinates of the region’s MBR with the four corners of the center of the location map. The cell number that is maximally covered by the region is supposedly to be assigned as the location index. However, the Sub-Block technique is not reliable because in most cases, the location index assigned is not the cell number that is maximally covered by the region and sometimes a region does not overlap with the cell number assigned at all. The effectiveness of this technique has been improved using the Improved Sub-Block technique by taking into consideration the total horizontal and vertical distances of a region at each location where it overlaps. The color-spatial technique is accessed on a Query-byExample CBIR system consisting of 900 images. From the experiments it is shown that retrieval effectiveness has been significantly improved by 85.86%.

https://ajba.um.edu.my/index.php/MJCS/article/download/6314/3992

Content-based image retrieval based on color-spatial features

Many fruit recognition approaches today are designed to classify different type of fruits but there is little effort being done for content-based fruit recognition specifically focuses on durian species. Durian, known as the king of tropical fruits, have few similar characteristics between different species where the skin have almost the same colour from green to yellowish brown with just slightly different shape and pattern of thorns. Therefore, it is hard to differentiate them with the current methods. It would be valuable to have an automated content-based recognition framework that can automatically represent and recognise a durian species given a durian image as the input. Therefore, this work aims to contribute to a new representation method based on multiple features for effective durian recognition. Two features based on shape and texture is considered in this work. Simple shape signatures which include area, perimeter, and circularity are used to determine the shape of the fruit durian and its base while the texture of the fruit is constructed based on Local Binary Pattern. We extracted these features from 240 durian images and trained this proposed method using few classifiers. Based on 10-fold cross validation, it is found that Logistic Regression, Gaussian Naive Bayesian, and Linear Discriminant Analysis classifiers performed equally well with 100% achievement of accuracy, precision, recall, and F1-score. We further tested the proposed algorithm on larger dataset which consisted of 42337 fruit images (64 various categories). Experimental results based on larger and more general dataset have shown that the proposed multiple features trained on Linear Discriminant Analysis classifier able to achieve 72.38% accuracy, 73% precision, 72% recall, and 72% F1-score.

Durian recognition based on multiple features and linear discriminant analysis

In the past decade, the popular Bag of Visual Words approach has been applied to many computer vision tasks, including image classification, video search, robot localization, and texture recognition. Unfortunately, most approaches use intensity features and discard color information, an important characteristic of any image that is motivated by human vision. Besides, if background colors are higher than foreground ones, Dominant Color Descriptor (DCD) retrieves images that contain similar background colors correctly. On the other hand, just color feature extraction is not sufficient for similar objects with different color descriptors (e.g. white dog vs. black dog). To solve these problems, a new Salient DCD (SDCD) color descriptor is proposed to extract foreground color and add semantic information into DCD based on the color distances and salient object extraction methods. Besides, a new fusion model is presented to fuse SDCD histogram and PHOW MSDSIFT histogram. Performance evaluation on several datasets proves that the new approach outperforms other existing, state-of-the-art methods.

Mas Rina Mustaffa

Papers

Texture classification and discrimination for region-based image retrieval

Ontology-Based Semantic Image Segmentation Using Mixture Models and Multiple CRFs.

Content-based image retrieval based on color-spatial features

Durian recognition based on multiple features and linear discriminant analysis

An effective fusion model for image retrieval