Plant Species Recognition Using Triangle-Distance Representation

ArXiv, 2018

This paper presents a novel feature set for shape-only leaf identification motivated by real-world, mobile deployment. The feature set includes basic shape features, as well as signal features extracted from local area integral invariants (LAIIs), similar to curvature maps, at multiple scales. The proposed methodology is evaluated on a number of publicly available leaf datasets with comparable results to existing methods which make use of colour and texture features in addition to shape. Over 90% classification accuracy is achieved on most datasets, with top-four accuracy for these datasets reaching over 98%. Rotation and scale invariance of the proposed features are demonstrated, along with an evaluation of the generalisability of the approach for generic shape matching.

2017

Plants are the most important living organisms on our planet because they are sources of energy and protect our planet against global warming. Botanists were the first scientist to design techniques for plant species recognition using leaves. Although many techniques for plant recognition using leaf images have been proposed in the literature, the precision and the quality of feature descriptors for shape, texture, and color remain the major challenges. This thesis investigates the precision of geometric shape features extraction and improved the determination of the Minimum Bounding Rectangle (MBR). The comparison of the proposed improved MBR determination method to Chaudhuri’s method is performed using Mean Absolute Error (MAE) generated by each method on each edge point of the MBR. On the top left point of the determined MBR, Chaudhuri’s method has the MAE value of 26.37 and the proposed method has the MAE value of 8.14. This thesis also investigates the use of the Convexity Meas...

Leaves can be categorized based on various physical attributes but the most popular attribute is shape. For example, we can use " shape " as a criterion to differentiate between a mango leaf and a banana leaf. Ideally this criterion can be of immense help to the user of an image database that stores, for instance, the images of different kinds of leaves. The user would often wish to retrieve images of leaf similar to the one he/she has or categorize the leaf according to its kind. This retrieval is mostly based upon the shape of the image. We investigate shape analysis methods for retrieving images and we develop an approach for doing so. In our approach we first find out the corners of the leaf by " Harris corner detection " and then determine the convex hull by joining these " corner points ". The leaves are then distinguished by finding the difference between their internal angles at each control point. The same criterion is then used to retrieve images of leaves from a plant database that are identical to the one in the input image.

In the frame of a tree species identifying mobile application, designed for a wide scope of users, and with didactic purposes, we developed a method based on the computation of explicit leaf shape descriptors inspired by the criteria used in botany. This paper focuses on the characterization of the leaf contour, the extraction of its properties, and its description using botanical terms. Contour properties are investigated using the Curvature-Scale Space representation, the potential teeth explicitly extracted and described, and the margin classified into a set of inferred shape classes. Results are presented for both margin shape characterization, and leaf classification over nearly 80 tree species.

Proceedings of the ACM International Conference on Image and Video Retrieval - CIVR '10, 2010

In this paper, an effective shape-based leaf image retrieval system is presented. A new contour descriptor is defined which reduces the number of points for the shape representation considerably. This shape representation is based on the curvature of the leaf contour and it deals with the scale factor in a novel and compact way. A two-step algorithm for retrieval is used. In a first step, the database is reduced using some geometrical features. Then a similarity measure between the contour representations is used to rank conveniently leaf images on the database. We implemented a prototype system based on these features and performed several experiments to show its effectiveness for plant species identification.

In this paper we compare two shape-based descriptors for plant leaf image classification. The leaves in the dataset are already segmented from the background only the contour detection algorithm is applied to extract the contour points and generate the shape-based descriptors. We propose a reduced size descriptor based on the angles between three points of the leaf contour and compare this descriptor with other similar descriptors based on their classification quality. The classification quality is measured both with 1-nearest neighbor comparison and with RBF-SVM model trained on the generated descriptors.

2013 IEEE International Conference on Multimedia and Expo (ICME), 2013

Unravelling mysteries of the diversity of the plant community is a crucial issue both for the development of many botanical industries as well as for the conservation of ecosystem biodiversity. Traditionally, botanists have proposed detailed dichotomous key descriptions (called also characters or concepts) about the morphology of plants and particularly of leaves that allow them to construct relationships between different plants and between plants and their environment. However, extracting these concepts is complicated, painstaking and can only be carried out by experts. One way to accelerate and broaden the use of these concepts is to automatically extract them directly from images. In this paper, we focus on one of the most basic and important concepts: the leaf arrangement. According to this concept, leaves are divided into four categories: simple, pinnnately compound, palmately compound and compound trifoliate. To accomplish this task, we follow a hierarchical scheme, reducing ambiguity between categories from the most different shapes to the most similar ones. The choice of appropriate features is performed based on botanical observations and validated by a statistical study. The method was tested on real world leaf images (the Pl@ntLeaves scans). Experimental results show its robustness for a high number of leaf species (70 species) and even in the presence of some distortions (such as rotation and partial leaf overlapping).

This paper proposes an automated system for recognizing plant species based on leaf images. Plant leaf images corresponding to three plant types, are analyzed using three different shape modelling techniques, the first two based on the Moments-Invariant (M-I) model and the Centroid-Radii (C-R) model and the third based on a proposed technique of Binary-Superposition (B-S). For the M-I model the first four central normalized moments have been considered. For the C-R model an edge detector has been used to identify the boundary of the leaf shape and 36 radii at 10 degree angular separation have been used to build the shape vector. The proposed approach consists of comparing binary versions of the leaf images through superposition and using the sum of non-zero pixel values of the resultant as the feature vector. The data set for experimentations consists of 180 images divided into training and testing sets and comparison between them is done using Manhattan, Euclidean and intersection norms. Accuracies obtained using the proposed technique is seen to be an improvement over the M-I and C-R based techniques, and comparable to the best figures reported in extant literature.

Fig. 1. Leaf shape modeling in the frequency domain. This figure shows the digital image process in combination with the Fourier descriptor for extracting a leaf shape signature. From left to right: original image; binary representation; extracted leaf contour; edge-centroid distance measure; normalised Fourier coefficients. Abstract-Plant identification and classification play an important role in ecology, but the manual process is cumbersome even for experimented taxonomists. Technological advances allows the development of strategies to make these tasks easily and faster. In this context, this paper describes a methodology for plant identification and classification based on leaf shapes, that explores the discriminative power of the contour-centroid distance in the Fourier frequency domain in which some invariance (e.g. rotation and scale) are guaranteed. In addition, it is also investigated the influence of feature selection techniques regarding classification accuracy. Our results show that by combining a set of features vectors-in the principal components space-and a feedforward neural network, an accuracy of 97.45% was achieved.

2013

A number of shape features for automatic plant recognition based on digital image processing have been proposed by Pauwels et al. in 2009. A database with 15 classes and 171 leaf samples was considered for the evaluation of these measures using linear discriminant analysis and hierarchical clustering. The results obtained match the human visual shape perception with an overall accuracy of 87%.