Jayme Garcia Arnal Barbedo

Digital image processing techniques for detecting, quantifying and classifying plant diseases

This paper presents a survey on methods that use digital image processing techniques to detect, quantify and classify plant diseases from digital images in the visible spectrum. Although disease symptoms can manifest in any part of the plant, only methods that explore visible symptoms in leaves and stems were considered. This was done for two main reasons: to limit the length of the paper and because methods dealing with roots, seeds and fruits have some peculiarities that would warrant a specific survey. The selected proposals are divided into three classes according to their objective: detection, severity quantification, and classification. Each of those classes, in turn, are subdivided according to the main technical solution used in the algorithm. This paper is expected to be useful to researchers working both on vegetable pathology and pattern recognition, providing a comprehensive and accessible overview of this important field of research.This paper presents a survey on methods that use digital image processing techniques to detect, quantify and classify plant diseases from digital images in the visible spectrum. Although disease symptoms can manifest in any part of the plant, only methods that explore visible symptoms in leaves and stems were considered. This was done for two main reasons: to limit the length of the paper and because methods dealing with roots, seeds and fruits have some peculiarities that would warrant a specific survey. The selected proposals are divided into three classes according to their objective: detection, severity quantification, and classification. Each of those classes, in turn, are subdivided according to the main technical solution used in the algorithm. This paper is expected to be useful to researchers working both on vegetable pathology and pattern recognition, providing a comprehensive and accessible overview of this important field of research.

A Review on Methods for Automatic Counting of Objects in Digital Images

A growing number of routine and research activities, in a wide variety of fields, have the counting of certain types of objects (cells, people, insects, etc.) as one of their main components. In most cases, such counting procedure is performed manually, in a process that is often lengthy and tedious. For that reason, several methods for automatically counting the objects of interest have been proposed in the last two decades. The vast majority of those methods rely on digital images containing the objects to provide an estimate as close as possible to the results manually obtained by human experts. The review is organized in a tutorial-like form, that is, instead of grouping the references according to a given criterion and then describing them, the paper describes some of the main tools and techniques used in this field of research, and then cites the references as sources for additional information and inspiration.

Method for Counting Microorganisms and Colonies in Microscopic Images

This paper presents a method to count microorganisms and colonies in microscopic images. The method uses a series of morphological operations to create a representation in which the objects of interest are easily isolated and counted. The proposal is successful in most cases, properly dealing with some difficult situations like when the sizes of the objects vary strongly and when there is low contrast between the objects and the background. Studies are underway in order to improve the performance of the method when dealing with strongly merged objects.

3D Plant Modeling: Localization, Mapping and Segmentation for Plant Phenotyping Using a Single Hand-held Camera

Functional-structural modeling and high-throughput phenomics demand tools for 3D measurements of plants. In this work, structure from motion is employed to estimate the position of a hand-held camera, moving around plants, and to recover a sparse 3D point cloud sampling the plants’ surfaces. Multiple-view stereo is employed to extend the sparse model to a dense 3D point cloud. The model is automatically segmented by spectral clustering, properly separating the plant’s leaves whose surfaces are estimated by fitting trimmed B-splines to their 3D points. These models are accurate snapshots for the aerial part of the plants at the image acquisition moment and allow the measurement of different features of the specimen phenotype. Such state-of-the-art computer vision techniques are able to produce accurate 3D models for plants using data from a single free moving camera, properly handling occlusions and diversity in size and structure for specimens presenting sparse canopies. A data set formed by the input images and the resulting camera poses and 3D points clouds is available, including data for sunflower and soybean specimens.

Survey on automatic transcription of music

An automatic music transcriber is a device that detects, without human interference, the musical gestures required to play a particular piece. Many techniques have been proposed to solve the problem of automatic music transcription. This paper presents an overview on the theme, discussing digital signal processing techniques, pattern classification techniques and heuristic assumptions derived from music knowledge that were used to build some of the main systems found in the literature. The paper is focused on the motivations behind each technique, aiming to serve both as an introduction to the theme and as resource for the development of new solutions for automatic transcription.

Method for Automatic Counting Root Nodules Using Digital Images

This paper presents a method to automatically count nodules that are present in the roots of many legume plants, using digital images captured after the nodules have been removed from the roots. This problem poses a significant challenge due to a number of reasons: the size and shape of the nodules vary greatly, they may appear clustered, and their texture is not uniform. The proposed method exploits well-known morphological operations to discriminate each nodule, making the counting process possible. Tests reveal a good correlation between automatic and manual counting.

A new automatic method for disease symptom segmentation in digital photographs of plant leaves

The segmentation of symptoms during image analysis of diseased plant leaves is an essential process for detection and classification of diseases. However, there are challenges involved in the task, many of them related to the variability of image and host/symptom characteristics and conditions. As a result of those challenges, the methods proposed in the literature so far focus on a specific problem and are usually bounded by tight constraints regarding image capture conditions. This research explores a new automatic method for segmenting disease symptoms on plant leaves that was designed to be applicable in a wide range of situations. The proposed technique employs only color channel manipulations and Boolean operations applied on binary masks, thus being simpler and more robust compared to many previously described automatic methods. Its effectiveness is demonstrated by tests performed over a large database containing images of 77 different diseases of 11 plant species. A comparison with manual segmentation is also presented, further reinforcing the advantages of the proposed approach.

Expert Systems Applied to Plant Disease Diagnosis: Survey and Critical View

Expert systems have been applied to solve agricultural problems for some time. The complexities involved in plant disease diagnosis make this problem a prime candidate for the application of expert systems. Those same complexities, however, make the development of a viable system a very challenging task. Many attempts at creating such tools have been reported in the literature, with different scopes and degrees of success. This article had the objective of organizing all the information that has been published on the subject in the last four decades by presenting a brief individual description of each proposed system, and then condensing all the information into a critical analysis of how the problem was handled in the past, how it evolved into the current scenario, and which are the possible directions to be explored in the future.

Counting Clustered Soybean Seeds

This paper presents a method to automatically count clustered soybean seeds using digital images. The method is based on classical morphological operations, and was designed to deal with the main difficulties imposed by images of soybean seeds, namely the clustering of the seeds, variations in the illumination, and low contrast between seeds and background. The proposal shows a good performance under a wide variety of condition.

Impact of dataset size and variety on the effectiveness of deep learning and transfer learning for plant disease classification

Abstract The problem of automatic recognition of plant diseases has been historically based on conventional machine learning techniques such as Support Vector Machines, Multilayer Perceptron Neural Networks and Decision Trees. However, the prevailing approach has shifted to the application of deep learning concepts, with focus on Convolutional Neural Networks (CNNs). In general, this kind of technique requires large datasets containing a wide variety of conditions to work properly. This is an important limitation, given the many challenges involved in the construction of a suitable image database. In this context, this study investigates how the size and variety of the datasets impact the effectiveness of deep learning techniques applied to plant pathology. This investigation was based on an image database containing 12 plant species, each presenting very different characteristics in terms of number of samples, number of diseases and variety of conditions. Experimental results indicate that while the technical constraints linked to automatic plant disease classification have been largely overcome, the use of limited image datasets for training brings many undesirable consequences that still prevent the effective dissemination of this type of technology.