Núbia Rosa da Silva

Plant Identification Based on Leaf Midrib Cross-Section Images Using Fractal Descriptors

The correct identification of plants is a common necessity not only to researchers but also to the lay public. Recently, computational methods have been employed to facilitate this task, however, there are few studies front of the wide diversity of plants occurring in the world. This study proposes to analyse images obtained from cross-sections of leaf midrib using fractal descriptors. These descriptors are obtained from the fractal dimension of the object computed at a range of scales. In this way, they provide rich information regarding the spatial distribution of the analysed structure and, as a consequence, they measure the multiscale morphology of the object of interest. In Biology, such morphology is of great importance because it is related to evolutionary aspects and is successfully employed to characterize and discriminate among different biological structures. Here, the fractal descriptors are used to identify the species of plants based on the image of their leaves. A large number of samples are examined, being 606 leaf samples of 50 species from Brazilian flora. The results are compared to other imaging methods in the literature and demonstrate that fractal descriptors are precise and reliable in the taxonomic process of plant species identification.

Brachiaria species identification using imaging techniques based on fractal descriptors

Abstract The use of a rapid and accurate method in diagnosis and classification of species and/or cultivars of forage has practical relevance, scientific and trade in various areas of study, since it has broad representation in grazing from tropical regions. Nowadays it occupies about 90% of the grazing area along Brazil and, besides the grazing areas to feed ruminants, Brachiaria also corresponds to about 80% of seeds being traded in all the world, bringing a large amount of money to Brazil. To identify species and/or cultivars of this genus is of fundamental importance in the fields that produce seeds, to ensure varietal purity and the effectiveness of improvement programs. Thus, leaf samples of fodder plant species Brachiaria were previously identified, collected and scanned to be treated by means of artificial vision to make the database and be used in subsequent classifications. Forage crops used were: Brachiaria decumbens cv. IPEAN; Brachiaria ruziziensis Germain & Evrard; Brachiaria brizantha (Hochst. ex. A. Rich.) Stapf; Brachiaria arrecta (Hack.) Stent. and Brachiaria spp. The images were analyzed by the fractal descriptors method, where a set of measures are obtained from the values of the fractal dimension at different scales. Therefore such values are used as inputs for a state-of-the-art classifier, the Support Vector Machine, which finally discriminates the images according to the respective species. The proposed method outperforms other state-of-the-art image analysis methods and makes possible the correct prediction of species in more than 93% of the samples. Such remarkable result is consequence of the better suitability of representing complex structures like those arising in the plant leaves by measures of complexity from fractal geometry. Finally, this high correctness rate suggests that the fractal method is an important tool to help the botanist.

Improved texture image classification through the use of a corrosion-inspired cellular automaton

Abstract In this paper, the problem of classifying synthetic and natural texture images is addressed. To tackle this problem, an innovative method is proposed that combines concepts from corrosion modeling and cellular automata to generate a texture descriptor. The core processes of metal (pitting) corrosion are identified and applied to texture images by incorporating the basic mechanisms of corrosion in the transition function of the cellular automaton. The surface morphology of the image is analyzed before and during the application of the transition function of the cellular automaton. In each iteration the cumulative mass of corroded product is obtained to construct each of the attributes of the texture descriptor. In the final step, this texture descriptor is used for image classification by applying Linear Discriminant Analysis. The method was tested on the well-known Brodatz and Vistex databases. In addition, in order to verify the robustness of the method, its invariance to noise and rotation was tested. To that end, different variants of the original two databases were obtained through addition of noise to and rotation of the images. The results showed that the proposed texture descriptor is effective for texture classification according to the high success rates obtained in all cases. This indicates the potential of employing methods taking inspiration from natural phenomena in other fields.

Leaf epidermis images for robust identification of plants

This paper proposes a methodology for plant analysis and identification based on extracting texture features from microscopic images of leaf epidermis. All the experiments were carried out using 32 plant species with 309 epidermal samples captured by an optical microscope coupled to a digital camera. The results of the computational methods using texture features were compared to the conventional approach, where quantitative measurements of stomatal traits (density, length and width) were manually obtained. Epidermis image classification using texture has achieved a success rate of over 96%, while success rate was around 60% for quantitative measurements taken manually. Furthermore, we verified the robustness of our method accounting for natural phenotypic plasticity of stomata, analysing samples from the same species grown in different environments. Texture methods were robust even when considering phenotypic plasticity of stomatal traits with a decrease of 20% in the success rate, as quantitative measurements proved to be fully sensitive with a decrease of 77%. Results from the comparison between the computational approach and the conventional quantitative measurements lead us to discover how computational systems are advantageous and promising in terms of solving problems related to Botany, such as species identification.

Texture recognition based on diffusion in networks

A texture recognition method is proposed.The image is modeled into networks.The texture is characterized by the diffusion over the networks.It is demonstrated that directed networks are better to texture recognition.The proposed method outperform the state-of-the-art. Much work has been done in the field of texture analysis and classification. While promising classification methods have been proposed, most of them rely on classical image analysis approaches. This paper presents a texture classification method based on diffusion in directed networks. First, an image is modeled as a directed network by mapping each pixel as a node and connecting two nodes up to a maximum distance r. To reveal texture properties, links between two nodes are removed based on the pixel intensity difference. Once such a network is obtained, the activity of each node is estimated by random walks and combined into a histogram to describe the image. The main contribution of this paper is the use of directed networks, which tends to provide better performance than in undirected cases. Also, we have shown that the activity induced on these networks can be effectively used as texture descriptor. Experimental results show that the proposed method is favorably compared to traditional texture methods on widely used texture datasets. The proposed method is also found to be promising for plant species classification using samples of leaf texture.

Performing edge detection by Difference of Gaussians using q-Gaussian kernels

In image processing, edge detection is a valuable tool to perform the extraction of features from an image. This detection reduces the amount of information to be processed, since the redundant information (considered less relevant) can be disconsidered. The technique of edge detection consists of determining the points of a digital image whose intensity changes sharply. This changes are, for example, due to the discontinuities of the orientation on a surface. A well known method of edge detection is the Difference of Gaussians (DoG). The method consists of subtracting two Gaussians, where a kernel has a standard deviation smaller than the previous one. The convolution between the subtraction of kernels and the input image results in the edge detection of this image. This paper introduces a method of extracting edges using DoG with kernels based on the q-Gaussian probability distribution, derived from the q-statistic proposed by Constantino Tsallis. To demonstrate the methods potential, we compare the introduced method with the tradicional DoG using Gaussians kernels. The results showed that the proposed method can extract edges with more accurate details.

Automated classification of wood transverse cross-section micro-imagery from 77 commercial Central-African timber species

Abstract• Key messagePattern recognition has become an important tool to aid in the identification and classification of timber species. In this context, the focus of this work is the classification of wood species using texture characteristics of transverse cross-section images obtained by microscopy. The results show that this approach is robust and promising.• ContextConsidering the lack of automated methods for wood species classification, machine vision based on pattern recognition might offer a feasible and attractive solution because it is less dependent on expert knowledge, while existing databases containing high-quality microscopy images can be exploited.• AimsThis work focuses on the automated classification of 1221 micro-images originating from 77 commercial timber species from the Democratic Republic of Congo.• MethodsMicroscopic images of transverse cross-sections of all wood species are taken in a standardized way using a magnification of 25 ×. The images are represented as texture feature vectors extracted using local phase quantization or local binary patterns and classified by a nearest neighbor classifier according to a triplet of labels (species, genus, family).• ResultsThe classification combining both local phase quantization and linear discriminant analysis results in an average success rate of approximately 88% at species level, 89% at genus level and 90% at family level. The success rate of the classification method is remarkably high. More than 50% of the species are never misclassified or only once. The success rate is increasing from the species, over the genus to the family level. Quite often, pattern recognition results can be explained anatomically. Species with a high success rate show diagnostic features in the images used, whereas species with a low success rate often have distinctive anatomical features at other microscopic magnifications or orientations than those used in our approach.• ConclusionThis work demonstrates the potential of a semi-automated classification by resorting to pattern recognition. Semi-automated systems like this could become valuable tools complementing conventional wood identification.

Heterogeneous patterns enhancing static and dynamic texture classification

Some mixtures, such as colloids like milk, blood, and gelatin, have homogeneous appearance when viewed with the naked eye, however, to observe them at the nanoscale is possible to understand the heterogeneity of its components. The same phenomenon can occur in pattern recognition in which it is possible to see heterogeneous patterns in texture images. However, current methods of texture analysis can not adequately describe such heterogeneous patterns. Common methods used by researchers analyse the image information in a global way, taking all its features in an integrated manner. Furthermore, multi-scale analysis verifies the patterns at different scales, but still preserving the homogeneous analysis. On the other hand various methods use textons to represent the texture, breaking texture down into its smallest unit. To tackle this problem, we propose a method to identify texture patterns not small as textons at distinct scales enhancing the separability among different types of texture. We find sub patterns of texture according to the scale and then group similar patterns for a more refined analysis. Tests were performed in four static texture databases and one dynamical one. Results show that our method provide better classification rate compared with conventional approaches both in static and in dynamic texture.