Limited Image Dataset Research Articles

Weed Detection and Classification with Computer Vision Using a Limited Image Dataset

In agriculture, as precision farming increasingly employs robots to monitor crops, the use of weeding and harvesting robots is expanding the need for computer vision. Currently, most researchers and companies address these computer vision tasks with CNN-based deep learning. This technology requires large datasets of plant and weed images labeled by experts, as well as substantial computational resources. However, traditional feature-based approaches to computer vision can extract meaningful parameters and achieve comparably good classification results with only a tenth of the dataset size. This study presents these methods and seeks to determine the minimum number of training images required to achieve reliable classification. We tested the classification results with 5, 10, 20, 40, 80, and 160 images per weed type in a four-class classification system. We extracted shape features, distance transformation features, color histograms, and texture features. Each type of feature was tested individually and in various combinations to determine the best results. Using six types of classifiers, we achieved a 94.56% recall rate with 160 images per weed. Better results were obtained with more training images and a greater variety of features.

Perception without preconception: comparison between the human and machine learner in recognition of tissues from histological sections

Deep neural networks (DNNs) have shown success in image classification, with high accuracy in recognition of everyday objects. Performance of DNNs has traditionally been measured assuming human accuracy is perfect. In specific problem domains, however, human accuracy is less than perfect and a comparison between humans and machine learning (ML) models can be performed. In recognising everyday objects, humans have the advantage of a lifetime of experience, whereas DNN models are trained only with a limited image dataset. We have tried to compare performance of human learners and two DNN models on an image dataset which is novel to both, i.e. histological images. We thus aim to eliminate the advantage of prior experience that humans have over DNN models in image classification. Ten classes of tissues were randomly selected from the undergraduate first year histology curriculum of a Medical School in North India. Two machine learning (ML) models were developed based on the VGG16 (VML) and Inception V2 (IML) DNNs, using transfer learning, to produce a 10-class classifier. One thousand (1000) images belonging to the ten classes (i.e. 100 images from each class) were split into training (700) and validation (300) sets. After training, the VML and IML model achieved 85.67 and 89% accuracy on the validation set, respectively. The training set was also circulated to medical students (MS) of the college for a week. An online quiz, consisting of a random selection of 100 images from the validation set, was conducted on students (after obtaining informed consent) who volunteered for the study. 66 students participated in the quiz, providing 6557 responses. In addition, we prepared a set of 10 images which belonged to different classes of tissue, not present in training set (i.e. out of training scope or OTS images). A second quiz was conducted on medical students with OTS images, and the ML models were also run on these OTS images. The overall accuracy of MS in the first quiz was 55.14%. The two ML models were also run on the first quiz questionnaire, producing accuracy between 91 and 93%. The ML models scored more than 80% of medical students. Analysis of confusion matrices of both ML models and all medical students showed dissimilar error profiles. However, when comparing the subset of students who achieved similar accuracy as the ML models, the error profile was also similar. Recognition of ‘stomach’ proved difficult for both humans and ML models. In 04 images in the first quiz set, both VML model and medical students produced highly equivocal responses. Within these images, a pattern of bias was uncovered–the tendency of medical students to misclassify ‘liver’ tissue. The ‘stomach’ class proved most difficult for both MS and VML, producing 34.84% of all errors of MS, and 41.17% of all errors of VML model; however, the IML model committed most errors in recognising the ‘skin’ class (27.5% of all errors). Analysis of the convolution layers of the DNN outlined features in the original image which might have led to misclassification by the VML model. In OTS images, however, the medical students produced better overall score than both ML models, i.e. they successfully recognised patterns of similarity between tissues and could generalise their training to a novel dataset. Our findings suggest that within the scope of training, ML models perform better than 80% medical students with a distinct error profile. However, students who have reached accuracy close to the ML models, tend to replicate the error profile as that of the ML models. This suggests a degree of similarity between how machines and humans extract features from an image. If asked to recognise images outside the scope of training, humans perform better at recognising patterns and likeness between tissues. This suggests that ‘training’ is not the same as ‘learning’, and humans can extend their pattern-based learning to different domains outside of the training set.

Convolutional Neural Network-Based Skin Lesion Classification With Variable Nonlinear Activation Functions

One of the worst forms of skin cancer is melanoma which can be curable if it is diagnosed at an early stage. The earlier the cancer is diagnosed, the better is the outcome. The risk of death from melanoma is directly related to the delay in identifying a lesion. A Deep Learning-based computer diagnosing system can be an automated solution in clinical assessments to overcome this problem. Convolutional Neural Network (CNN) can help to improve the classification rate of skin lesions from dermoscopic images without the need for any human assistance. The linear and nonlinear activation functions act as a node placed at hidden layers or output layers of a Neural Network to play a role in deciding whether that node should pass information to the following layer or not. This critical mathematical mechanism influences the accuracy rate of the CNN. To obtain acceptable performance, CNN requires a large amount of training data. This research shows how fast and effective different types of nonlinear activation functions work on a CNN with limited image datasets. Experimental analysis reveals that the proposed CNN model with parameterized Leaky ReLU function outperforms (97.50% accuracy, 98.00% precision, and 98.00% sensitivity) the same network with distinct nonlinear activation functions for the problem of melanoma recognition by classifying skin lesion into three classes. All experimental studies are carried out using images from PH2 (a dermoscopic image database obtained at the Hospital Pedro Hispano Dermatology Service in Matosinhos, Portugal) and International Skin Imaging Collaboration (ISIC) archive datasets.

DEEP LEARNING AND STATISTICAL MODELS FOR DETECTION OF WHITE STEM BORER DISEASE IN ARABICA COFFEE

Abstract. Early detection of crop pest and disease is very critical for taking up suitable control measures to reduce the loss of economic yield. Coffee is an important commercial crop in India which is affected by pests and diseases every year resulting in major yield loss. White stem borer (Xylotrechus quadripes) is the most serious pest of coffee (Arabica sp.) in India causing substantial loss of yield every year. Detection of the infestation in its early stage is quite challenging. In this regard, image pattern recognition techniques offer cost effective and scalable solutions. An image library was created representing different stages of the plant infestation using camera/mobile devices. Our Convolutional Neural Network (CNN) models use these images of healthy and infested plants for early detection of white stem borer infestation. The overall methodology included image processing, machine learning, supervised transfer learning and unsupervised auto-encoding techniques to solve the problem of early detection and severity of the infestation. Using the Inception v3 transfer learning model, we obtained average accuracy of 85.5% which is quite encouraging with limited image datasets. We explore Unsupervised Autoencoder models, which can work with limited image datasets. In addition, statistical analysis of long-term climatic factors such as temperature, rainfall, humidity and luminescence is explored for reliable detection and diagnosis of the infestation. Based on the encouraging results, a mobile application is proposed for near real time monitoring of WSB infestation to help the coffee planter’s community.

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

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.

Deformable Image Registration based on Similarity-Steered CNN Regression.

Existing deformable registration methods require exhaustively iterative optimization, along with careful parameter tuning, to estimate the deformation field between images. Although some learning-based methods have been proposed for initiating deformation estimation, they are often template-specific and not flexible in practical use. In this paper, we propose a convolutional neural network (CNN) based regression model to directly learn the complex mapping from the input image pair (i.e., a pair of template and subject) to their corresponding deformation field. Specifically, our CNN architecture is designed in a patch-based manner to learn the complex mapping from the input patch pairs to their respective deformation field. First, the equalized active-points guided sampling strategy is introduced to facilitate accurate CNN model learning upon a limited image dataset. Then, the similarity-steered CNN architecture is designed, where we propose to add the auxiliary contextual cue, i.e., the similarity between input patches, to more directly guide the learning process. Experiments on different brain image datasets demonstrate promising registration performance based on our CNN model. Furthermore, it is found that the trained CNN model from one dataset can be successfully transferred to another dataset, although brain appearances across datasets are quite variable.

TH-D-AUD A-08: A Method for Generating Large Datasets of Patient Geometries for Radiotherapy Treatment Planning Studies

Purpose: Radiotherapy treatment planning studies often require the use of large patient CT datasets to extract conclusions of statistical significance. However, due to various reasons such as difficulties in acquiring a larger set of CT scans and in segmenting organs in all the images, real studies are usually performed using very limited datasets. The aim of this study is to develop a novel method for generating large datasets of realistic patient geometries for treatment planning studies using a computerized phantom. Method and Materials: A NURBS-based cardiac-torso (NCAT) phantom was built based upon data from the female Chinese Visible Human (CVH) datasets. The NURBS control points on the organs surface were deformed to match the organ surface obtained from the limited daily cone beam CT (CBCT) dataset of each patient undergoing adaptive radiotherapy. A principal component analysis (PCA) of control point deformations was performed for the individual patient and each geometry was fit to a combination of the principal components with their corresponding weighting factors. A statistical analysis of the weighting factors was performed, and a new larger set of statistically equivalent weighting factors can be constructed, which will result in a larger geometry dataset for the patient. Results: The NCAT pelvis phantom was developed based on the segmentation of organs in the Chinese Visible Human pelvic dataset. The method for generating new datasets was applied to 20 patients undergoing adaptive radiotherapy and a variety of realistic deformed pelvic geometries was developed. Conclusion: We present a novel method for automatic generation of large datasets of patient geometries from a set of limited image datasets. The new geometries should have the same statistical uncertainties as the original datasets, however much smaller random uncertainties, and can be used in the future for adaptive radiotherapy planning studies.