Automatic text block separation in document images

Separation of printed text blocks from the non-text areas, containing signatures, handwritten text, logos and other such symbols, is a necessary first step for an OCR involving printed text recognition. In the present work, we compare the efficacy of some feature-classifier combinations to carry out this separation task. We have selected length-nomalized horizontal projection profile (HPP) as the starting point of such a separation task. This is with the assumption that the printed text blocks contain lines of text which generate HPP's with some regularity. Such an assumption is demonstrated to be valid. Our features are the HPP and its two transformed versions, namely, eigen and Fisher profiles. Four well known classifiers, namely, Nearest neighbor, Linear discriminant function, SVM's and artificial neural networks have been considered and efficiency of the combination of these classifiers with the above features is compared. A sequential floating feature selection technique has been adopted to enhance the efficiency of this separation task. The results give an average accuracy of about 96.

Tibetan Written Images : A Study of Imagery in the Writings of Dhondup Gyal

Dhondup Gyal (Don grub rgyal, 1953 - 1985) was a Tibetan writer from Amdo (Qinghai, People's Republic of China). He wrote several prose works, poems, scholarly writings and other works which have been later on collected together into The Collected Works of Dhondup Gyal, in six volumes. He had a remarkable influence on the development of modern Tibetan literature in the 1980s. Examining his works, which are characterized by rich imagery, it is possible to notice a transition from traditional to modern ways of literary expression. Imagery is found in both the poems and prose works of Dhondup Gyal. Nature imagery is especially prominent and his writings contain images of flowers and plants, animals, water, wind and clouds, the heavenly bodies and other environmental elements. Also there are images of parts of the body and material and cultural images. To analyse the images, most of which are metaphors and similes, the use of the cognitive theory of metaphor provides a good framework for making comparisons with images in traditional Tibetan literature and also some images in Chinese, Indian and Western literary works. The analysis shows that the images have both traditional and innovative features. The source domains of images often appear similar to those found in traditional Tibetan literature and are slow to change. However, innovative shifts occur in the way they are mapped on their target domains, which may express new meanings and are usually secular in nature if compared to the religiosity which often characterizes traditional Tibetan literature. Dhondup Gyal's poems are written in a variety of styles, ranging from traditional types of verse compositions and poems in the ornate kāvya-style to modern free verse poetry. The powerful central images of his free verse poems and some other works can be viewed as structurally innovative and have been analysed with the help of the theory of conceptual blending. They are often ambiguous in their meaning, but can be interpreted to express ideas related to creativity, freedom and the need for change and development.

A Huber-loss-driven clustering technique and its application to robust cell detection in confocal microscopy images

We address the problem of detecting cells in biological images. The problem is important in many automated image analysis applications. We identify the problem as one of clustering and formulate it within the framework of robust estimation using loss functions. We show how suitable loss functions may be chosen based on a priori knowledge of the noise distribution. Specifically, in the context of biological images, since the measurement noise is not Gaussian, quadratic loss functions yield suboptimal results. We show that by incorporating the Huber loss function, cells can be detected robustly and accurately. To initialize the algorithm, we also propose a seed selection approach. Simulation results show that Huber loss exhibits better performance compared with some standard loss functions. We also provide experimental results on confocal images of yeast cells. The proposed technique exhibits good detection performance even when the signal-to-noise ratio is low.

Exploring Transfer Learning Approaches for Head Pose Classification from Multi-view Surveillance Images

Head pose classification from surveillance images acquired with distant, large field-of-view cameras is difficult as faces are captured at low-resolution and have a blurred appearance. Domain adaptation approaches are useful for transferring knowledge from the training (source) to the test (target) data when they have different attributes, minimizing target data labeling efforts in the process. This paper examines the use of transfer learning for efficient multi-view head pose classification with minimal target training data under three challenging situations: (i) where the range of head poses in the source and target images is different, (ii) where source images capture a stationary person while target images capture a moving person whose facial appearance varies under motion due to changing perspective, scale and (iii) a combination of (i) and (ii). On the whole, the presented methods represent novel transfer learning solutions employed in the context of multi-view head pose classification. We demonstrate that the proposed solutions considerably outperform the state-of-the-art through extensive experimental validation. Finally, the DPOSE dataset compiled for benchmarking head pose classification performance with moving persons, and to aid behavioral understanding applications is presented in this work.

Low Resolution Face Recognition Across Variations in Pose and Illumination

We propose a completely automatic approach for recognizing low resolution face images captured in uncontrolled environment. The approach uses multidimensional scaling to learn a common transformation matrix for the entire face which simultaneously transforms the facial features of the low resolution and the high resolution training images such that the distance between them approximates the distance had both the images been captured under the same controlled imaging conditions. Stereo matching cost is used to obtain the similarity of two images in the transformed space. Though this gives very good recognition performance, the time taken for computing the stereo matching cost is significant. To overcome this limitation, we propose a reference-based approach in which each face image is represented by its stereo matching cost from a few reference images. Experimental evaluation on the real world challenging databases and comparison with the state-of-the-art super-resolution, classifier based and cross modal synthesis techniques show the effectiveness of the proposed algorithm.

Desenho-escrita a ponto de verbo

A pesquisa é um dos desdobramentos possíveis da ação do desenho como exigência do próprio corpo. O trabalho propõe a tentativa e o fracasso como movimentação do fazer artístico. Por acreditar na experiência cotidiana com a força escritural, em seu caráter performático, investe na possibilidade de ativar o desenho latente no mundo. Ao investigar a palavra como tentativa de ordenação e como uma condição de possibilidade da arte contemporânea, discute-se (e, simultaneamente, pratica-se) o risco, o acidente introduzido pelo poético em seu caráter desestabilizador,a fim de debruçar-se sobre o aspecto não estável das relações prático-teóricas, entre o texto e a matéria plástica, entre imagens e palavras. A dissertação busca, (na tensão da) experiência, a reflexão inesgotável diante da energia gasta na convivência com a matéria-pensamento, com a palavra que é carne. O horizonte da pesquisa é a impossibilidade - dentro do texto do artista que tangencia seu próprio trabalho de colocar-se em movimento em direção a um fim que não seja o caminho.

Trânsitos poéticos no ensino da arte: as visualidades, subjetividades e afetividades no cotidiano escolar

A pesquisa tem como foco a discussão das visualidades encontradas e, principalmente, produzidas no/do/com o cotidiano escolar, e suas relações com os currículos. O estudo se desenvolveu, tomando como base, o registro e recolhimento de dados a partir da nossa prática docente no ensino de arte, numa escola municipal do Rio de Janeiro. A escolha da problemática deste estudo surgiu do desejo de discutirmos as experiências vividas cotidianamente nos espaçostempos da escola, e de refletirmos em torno dos posicionamentos dos sujeitos deste contexto, frente às fissuras epistemológicas e conceituais que aí se apresentam. O debate em torno dos currículos vigentes e os praticados estão inseridos nesta investigação, trazendo à tona as imposições oficiais e as tessituras captadas nos entrelinhas dos cotidianos. As imagens que cercam este espaço, tanto as criadas quanto as que invadem tal cenário, são discutidas aqui quanto ao seu potencial emancipatório. Abordamos os desafios encontrados no exercício do ensinar/aprender na perspectiva contemporânea, levando em consideração as múltiplas identidades que transitam poeticamente na escola e as subjetividades que as caracterizam. Buscamos, ainda, incentivar as relações dialógicas com os docentes e suas atuações no campo escolar, acreditando que é na prática, e na reflexão sobre ela, que nos formamos enquanto educadores

Arte & Educação em cultura visual: ensino, sentido, presença e invasões das imagens nos cotidianos

Parte da hipótese que a obra poética, artística ou não, tem força para além da mediação da palavra, ou seja, da afirmação de sentidos que obliterariam a eloquência da presença. Busca discorrer sobre a teoria da presença e sua importância na interlocução com a produção poética nas artes visuais contemporâneas. Aponta a utilidade dessa argumentação como perspectiva para problematizar a Cultura Visual e defender o investimento na elucidação do universo das imagens visuais como elemento de formação humana em sintonia com as questões da alteridade e com os tempos de hoje. Entende que o estudo equalizador entre as imagens visuais e as obras de arte visual favorece a autonomia dos indivíduos e o melhor aproveitamento do mundo das artes com menor risco de sujeição às hegemonias culturais

Pothole detection in asphalt pavement images

Pavement condition assessment is essential when developing road network maintenance programs. In practice, the data collection process is to a large extent automated. However, pavement distress detection (cracks, potholes, etc.) is mostly performed manually, which is labor-intensive and time-consuming. Existing methods either rely on complete 3D surface reconstruction, which comes along with high equipment and computation costs, or make use of acceleration data, which can only provide preliminary and rough condition surveys. In this paper we present a method for automated pothole detection in asphalt pavement images. In the proposed method an image is first segmented into defect and non-defect regions using histogram shape-based thresholding. Based on the geometric properties of a defect region the potential pothole shape is approximated utilizing morphological thinning and elliptic regression. Subsequently, the texture inside a potential defect shape is extracted and compared with the texture of the surrounding non-defect pavement in order to determine if the region of interest represents an actual pothole. This methodology has been implemented in a MATLAB prototype, trained and tested on 120 pavement images. The results show that this method can detect potholes in asphalt pavement images with reasonable accuracy.