Improved wood species identification based on multi-view imagery of the three anatomical planes

BACKGROUND

The identification of tropical African wood species based on microscopic imagery is a challenging problem due to the heterogeneous nature of the composition of wood combined with the vast number of candidate species. Image classification methods that rely on machine learning can facilitate this identification, provided that sufficient training material is available. Despite the fact that the three main anatomical sections contain information that is relevant for species identification, current methods only rely on transverse sections. Additionally, commonly used procedures for evaluating the performance of these methods neglect the fact that multiple images often originate from the same tree, leading to an overly optimistic estimate of the performance.

RESULTS

We introduce a new image dataset containing microscopic images of the three main anatomical sections of 77 Congolese wood species. A dedicated multi-view image classification method is developed and obtains an accuracy (computed using the naive but common approach) of 95%, outperforming the single-view methods by a large margin. An in-depth analysis shows that naive accuracy estimates can lead to a dramatic over-prediction, of up to 60%, of the accuracy.

CONCLUSIONS

Additional images from non-transverse sections can boost the performance of machine-learning-based wood species identification methods. Additionally, care should be taken when evaluating the performance of machine-learning-based wood species identification methods to avoid an overestimation of the performance.

Automatic Leaf Epidermis Assessment Using Fourier Descriptors in Texture Images

The identification of plant species is not a trivial task, since it is carried out by experts and depends on the presence of fruits, flowers and leaves. However, fruits and flowers are not available throughout the year, while leaves are accessible most of the year. In order to assist the specialized work of species identification, methods of texture image analysis are used to extract characteristics from samples of imaged leaves and thus predict the species. Texture image analysis is a versatile and powerful technique able to extract measurements from patterns in the images. Using this technique, recent research has found a close relationship between texture and plant species (da Silva et al., 2015 and 2016). Here, we describe the procedure to extract texture features from microscopic images of leaves using Fourier (Cosgriff, 1960; Azencott, 1997; Gonzalez and Woods, 2006). It is important to highlight that other methods for texture extraction can be used as well. This protocol is split into two parts: (A) leaf epidermal dissociation and (B) automatic method for leaf epidermal image analysis.

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.

Medical image retrieval and analysis by Markov random fields and multi-scale fractal dimension

Many Content-based Image Retrieval (CBIR) systems and image analysis tools employ color, shape and texture (in a combined fashion or not) as attributes, or signatures, to retrieve images from databases or to perform image analysis in general. Among these attributes, texture has turned out to be the most relevant, as it allows the identification of a larger number of images of a different nature. This paper introduces a novel signature which can be used for image analysis and retrieval. It combines texture with complexity extracted from objects within the images. The approach consists of a texture segmentation step, modeled as a Markov Random Field process, followed by the estimation of the complexity of each computed region. The complexity is given by a Multi-scale Fractal Dimension. Experiments have been conducted using an MRI database in both pattern recognition and image retrieval contexts. The results show the accuracy of the proposed method in comparison with other traditional texture descriptors and also indicate how the performance changes as the level of complexity is altered.

A systematic comparison of supervised classifiers

Pattern recognition has been employed in a myriad of industrial, commercial and academic applications. Many techniques have been devised to tackle such a diversity of applications. Despite the long tradition of pattern recognition research, there is no technique that yields the best classification in all scenarios. Therefore, as many techniques as possible should be considered in high accuracy applications. Typical related works either focus on the performance of a given algorithm or compare various classification methods. In many occasions, however, researchers who are not experts in the field of machine learning have to deal with practical classification tasks without an in-depth knowledge about the underlying parameters. Actually, the adequate choice of classifiers and parameters in such practical circumstances constitutes a long-standing problem and is one of the subjects of the current paper. We carried out a performance study of nine well-known classifiers implemented in the Weka framework and compared the influence of the parameter configurations on the accuracy. The default configuration of parameters in Weka was found to provide near optimal performance for most cases, not including methods such as the support vector machine (SVM). In addition, the k-nearest neighbor method frequently allowed the best accuracy. In certain conditions, it was possible to improve the quality of SVM by more than 20% with respect to their default parameter configuration.

Complex network classification using partially self-avoiding deterministic walks

Complex networks have attracted increasing interest from various fields of science. It has been demonstrated that each complex network model presents specific topological structures which characterize its connectivity and dynamics. Complex network classification relies on the use of representative measurements that describe topological structures. Although there are a large number of measurements, most of them are correlated. To overcome this limitation, this paper presents a new measurement for complex network classification based on partially self-avoiding walks. We validate the measurement on a data set composed by 40000 complex networks of four well-known models. Our results indicate that the proposed measurement improves correct classification of networks compared to the traditional ones.

Fractal descriptors in the Fourier domain applied to color texture analysis

The present work proposes the development of a novel method to provide descriptors for colored texture images. The method consists of two steps. First, we apply a linear transform in the color space of the image aiming at highlighting spatial structuring relations among the color of pixels. Second, we apply a multiscale approach to the calculus of fractal dimension based on Fourier transform. From this multiscale operation, we extract the descriptors that are used to discriminate the texture represented in digital images. The accuracy of the method is verified in the classification of two color texture datasets, by comparing the performance of the proposed technique to other classical and state-of-the-art methods for color texture analysis. The results showed an advantage of almost 3% of the proposed technique over the second best approach.

Mumford-Shah algorithm applied to videokeratography image processing and consequences to refractive power values

There are many corneal diseases that can be detected using an eye-care device called videokeratograph. The videokeratograph is based on the principle of an apparatus called Placido disc and is used to precisely measure the anterior surface of the cornea. This disc contains rings alternately white and black, which are reflected on the patient's cornea during the examination. The device can find anomalies by analyzing the reflected image, using image-processing algorithms. Although the efficiency of most commercial videokeratographs is acceptable, manufacturers do not disseminate to the scientific community the technique used in the image analysis algorithms. This makes it difficult for the specialized researcher in order to find better algorithms for the image-processing and, consequently, increase the instrument's precision. In this work we have segmented the Placido disc in polar coordinates by implementing a diagonal section of the image, in the radial direction. The objective is to find the inflection points in the signal obtained. In this paper the signal is studied by using the Mumford-Shah segmentation method. The results are compared to those obtained with other classic methods in the literature, e.g. Marr-Hildreth filters, numerical derivative, Fourier derivative, morphological Laplacian and Canny derivative. The best result was achieved by using the Mumford-Shah functional. Using this technique it was possible to find the inflection positions with higher accuracy. The method did not detect any false inflection. Mumford-Shah's method demonstrated also a high precision in the task of eliminating noises from the original signal.

[Different techniques for Plácido image analysis may improve precision of videokeratography]

PURPOSE

Videokeratography (VK) has been a widespread technology for corneal surface analysis since the mid-80s. The objective of this study was to develop different techniques for Plácido image edge detection and compare the results of each algorithm in terms of the consequences for axial curvature computations.

METHODS

Plácido images from an Eyesys system 2000 were captured for 4 different spherical surfaces. Each image was saved in bitmap format at the hard disk of an IBM computer. Six different image-processing algorithms were developed using different techniques well-documented in the literature. The six methods were as follows: (1) First order numerical derivative, (2) First and (3) Second order Fourier derivative, (4) the Marr-Hildreth filter, (5) Canny's Method, (6) Mathematical morphology. Each algorithm was tested on each of the Plácido images.

RESULTS

Edge radial distance from center of Plácido image was compared for each algorithm and a computer simulation of the videokeratography system. Mean deviation in terms of pixels/millimeters/dioptric power for all spheres for methods (1)-(6) were, respectively: (1) 33.1695/0.7961/0.79, (2) 32.79/0.7870/0.7724, (3) 60.7150/1.4572/1.4192, (4)18.97/0.4553/0.4572, (5) 46.33/1.1119/1.0917, (6) 20.55/0.4932/0.48.

CONCLUSION

Researchers and clinical ophthalmologists should be more careful when choosing commercial videokeratographs and also when comparing measurements of different instruments, given that there may be differences associated with the image processing technique. We have shown here that the Marr/Hildreth (method (4)) image processing method is more precise than other methods such as Fourier or first order numerical methods.

Spatial and frequency domain techniques for segmentation of Placido images and accuracy implications for videokeratography

OBJECTIVE

Videokeratography (VK) has been a widespread technology for corneal surface analysis since the mid 1980s. Most manufactures use personal computers attached to a Placido disc apparatus in order to capture and process digital images. Although precision reported by most manufactures are within very good limits, none of them have disclosed, probably due to proprietary reasons, the nature of the algorithm used in their image-processing phase. This is a problem when researchers want to reproduce or test their own curvature or elevation algorithms on Placido images generated on different commercial videokeratographs or even compare their algorithms on data from different manufactures. Our main objective in this work was to develop certain basic techniques for Placido image edge detection and to compare the results of each algorithm in terms of precision at the image level and also the consequences for axial curvature computations. We also propose that manufactures come forward and at least explain which image-processing technique is used in their own algorithms so other researchers and laboratories can make better use of their data to improve VK algorithms.

MATERIAL/METHODS

Placido images from an Eyesys system 2000 were captured for four different spherical surfaces. Each image was saved in bitmap format at the hard disk of an IBM computer. Six different image-processing algorithms were developed using different techniques well documented in the literature. The six methods were as follows: (1) first order numerical derivative, (2) first and (3) second order Fourier derivative, (4) the Marr-Hildreth filter, (5) Canny's method and (6) Mathematical Morphology. Each algorithm was tested on each of the Placido images.

RESULTS

Edge radial distance from center of Placido image was compared for each algorithm and a computer simulation of the VK system. The simulated image was used as absolute reference. Another approach was to calculate Axial dioptric power using, again, well documented procedures, and compare the results for each image detection algorithm. Mean deviation in terms of pixels/millimeters/dioptric power for all spheres for methods (1-6) were, respectively, (1) 33.1695/0.7961/0.79, (2) 32.79/0.7870/0.7724, (3) 60.7150/1.4572/1.4192, (4)18.97/0.4553/0.4572, (5) 46.33/1.1119/1.0917 and (6) 20.55/0.4932/0.48.

DISCUSSION

All methods have great deviation propagation in terms of dioptric power calculations when the axial algorithm is used and the absolute reference simulated edges are used to generate the calibration curves. This indicates that researchers should be more careful when using resulting image processing files from different videokeratographs to compare their own curvature or elevation algorithms among different instruments or even to measure the absolute precision of their new algorithms.

Leaf shape analysis using the multiscale Minkowski fractal dimension, a new morphometric method: a study with Passiflora (Passifloraceae)

A new method is proposed for the extraction of morphometric characteristics of plant leaf structures. A sample of 10 species of the genus Passiflora (P. coriacea Juss., P. foetida L., P. miersii Mast., P. organensis Gardner, P. pohlii Mast., P. suberosa L., P. amethystina J.C. Mikan, P. caerulea L., P. gibertii N.E.Br., P. maliformis L.) was used in an experiment to test the method. This genus shows a wide range of leaf forms, but there are some species pairs or groups whose morphological similarity makes their correct identification difficult. The multiscale function of the Minkowski fractal dimension was applied to digital images of leaves to generate complexity measures of their internal (veins) and external (leaf outline) form. The results of the leaf characteristic extraction method, as well as its potential as the basis for an identification mechanism, are discussed for the 10 species. The method was very accurate in correctly differentiating among species, since no leaf was erroneously identified. A small number of leaves per species was sufficient for establishing a characteristic pattern for each of them, which constitutes an important advantage of the method in the recognition and classification procedure.Key words: image processing, fractal dimension, plant taxonomy, morphometry, Passiflora.