Image Classification Techniques Research Articles

Methodological Ambiguity and Inconsistency Constrain Unmanned Aerial Vehicles as A Silver Bullet for Monitoring Ecological Restoration

The last decade has seen an exponential increase in the application of unmanned aerial vehicles (UAVs) to ecological monitoring research, though with little standardisation or comparability in methodological approaches and research aims. We reviewed the international peer-reviewed literature in order to explore the potential limitations on the feasibility of UAV-use in the monitoring of ecological restoration, and examined how they might be mitigated to maximise the quality, reliability and comparability of UAV-generated data. We found little evidence of translational research applying UAV-based approaches to ecological restoration, with less than 7% of 2133 published UAV monitoring studies centred around ecological restoration. Of the 48 studies, > 65% had been published in the three years preceding this study. Where studies utilised UAVs for rehabilitation or restoration applications, there was a strong propensity for single-sensor monitoring using commercially available RPAs fitted with the modest-resolution RGB sensors available. There was a strong positive correlation between the use of complex and expensive sensors (e.g., LiDAR, thermal cameras, hyperspectral sensors) and the complexity of chosen image classification techniques (e.g., machine learning), suggesting that cost remains a primary constraint to the wide application of multiple or complex sensors in UAV-based research. We propose that if UAV-acquired data are to represent the future of ecological monitoring, research requires a) consistency in the proven application of different platforms and sensors to the monitoring of target landforms, organisms and ecosystems, underpinned by clearly articulated monitoring goals and outcomes; b) optimization of data analysis techniques and the manner in which data are reported, undertaken in cross-disciplinary partnership with fields such as bioinformatics and machine learning; and c) the development of sound, reasonable and multi-laterally homogenous regulatory and policy framework supporting the application of UAVs to the large-scale and potentially trans-disciplinary ecological applications of the future.

Pre-processing spectrogram parameters improve the accuracy of bioacoustic classification using convolutional neural networks

ABSTRACTA variety of automated classification approaches have been developed to extract species detection information from large bioacoustic datasets. Convolutional neural networks (CNNs) are an image classification technique that can be operated on the spectrogram of an audio recording. Using CNNs for bioacoustic classification negates the need for sophisticated feature extraction techniques; however, CNNs may be sensitive to the parameters used to create spectrograms. We used AlexNet to classify spectrograms of audio clips from 19 species of birdsong. We trained and tested AlexNet with the spectrograms and observed that mean classification accuracy ranged from 88.9% to 96.9% depending on the parameters used to create the spectrogram. Classification accuracy was highest when we used a composite of four spectrograms with different combinations of scales for frequency and amplitude. Classification accuracy also varied depending on the FFT window size of the spectrogram. Overall, our results suggest that optimal spectrogram parameters for CNN classification may differ from those used for human visualization or other classification approaches. We suggest that if spectrogram parameters are appropriately selected, classification accuracy similar to current state-of-the-art methods can be achieved using off-the-shelf software and without the need to extract domain-specific features.

Comparative Study of Artificial Intelligence Techniques for Image Classification

There are various applications of image processing in the field of engineering, agriculture, graphic design, commerce, historical research and architecture. This paper studies and compares most of the research works done in the field of image processing and machine learning for the purpose of image classification based on the features extracted from the image through different feature extraction techniques. The machine learning techniques studied in this paper are Convolution Neural Network (CNN), Support Vector Machine (SVM) and Fuzzy logic. The paper studies and compares these methods for their implementation in classification of digital images. Color based segmentation models are used to segment the specific features from image and categories them into different classes. First image preprocessing is done on the image to reduce the noise from the image. Then image segmentation and edge detection techniques are used to identify the objects in the image and extract the features through which the image can be labeled with a specific class.

The achievement of higher flexibility in multiple-choice-based tests using image classification techniques

In spite of the high accuracy of the existing optical mark reading (OMR) systems and devices, a few restrictions remain existent. In this work, we aim to reduce the restrictions of multiple-choice questions (MCQ) within tests. We use an image registration technique to extract the answer boxes from answer sheets. Unlike other systems that rely on simple image processing steps to recognize the extracted answer boxes, we address the problem from another perspective by training a machine learning classifier to recognize the class of each answer box (i.e., confirmed, crossed out or blank answer). This gives us the ability to deal with a variety of shading and mark patterns, and distinguish between chosen (i.e., confirmed) and canceled answers (i.e., crossed out). All existing machine learning techniques require a large number of examples in order to train a model for classification; therefore, we present a dataset including six real MCQ assessments with different answer sheet templates. We evaluate two strategies of classification: a straightforward approach and a two-stage classifier approach. We test two handcrafted feature methods and a convolutional neural network. At the end, we present an easy-to-use graphical user interface of the proposed system. Compared with existing OMR systems, the proposed system has the least constraints and achieves a high accuracy. We believe that the presented work will further direct the development of OMR systems toward reducing the restrictions of the MCQ tests.

Classification of Brain Tumors Using Deep Features Extracted Using CNN

Deep learning play a major role in medical automation, Convolutional Neural Networks (CNN) is an important machine learning technique for medical image segmentation and classification. In this work, we propose a novel approach that uses CNN_S for classifying brain tumors in to normal and three different types. The tumor is initially segmented from the MRI images using an enhanced ICA mixture mode model. From the segmented image deep features are extracted and classified. The results have been validated by measuring the performance of classifier on a data set available with Harvard Medical School.

Assessment of green space cooling effects in dense urban landscape: a case study of Delhi, India

Most noticeable and direct impacts of urbanization on the environment is the changes in daily and seasonal thermal variable and increase the hydraulic stress. Both are cumulatively formed the urban heat island (UHI) effects. From the sustainability point of view, it is very crucial for research to focus on how to mitigate UHI. Therefore, objective of this study is to evaluate the cooling effects of different types of urban greenery at local and park level. For land surface temperature, Landsat 8—TIRS and for greenery mapping Sentinel 2A imageries have been used. LST was derived using the single channel algorithm, and greenery was mapping using combination of unsupervised (i.e., ISODATA) and supervised (i.e., maximum likelihood) image classification techniques. Numbers of class level metrics like PLAND PD; LPI; AREA_MN etc. used to measure composition and configuration of urban greenery. Ordinary least square regression, multiple linear regression, and spatial auto-regression model were used to measure the cooling effects of urban green space quantitatively. Besides, park level analysis also was done using park cooling intensity (PCI) and its relationship with park land use. Results shows that amount of urban greenery in terms of percentage of land cover or average size of the green patch is very important to reduce the UHI. PLAND, LPI, AREA_MN and COHESION negatively related with LST whereas PD, LSI, SHAPE_MN, and ENN_MN positively related with LST. So, fragmented vegetation area, green space with complex shape could increase the LST. If we look at the vegetation types and relation with LST, only dense vegetation and trees and plantation can effectively reduce the LST. Besides PCI negatively related with amount of built-up area and open space within parks. Finally, this study can help to the urban and land use planning to build a heat resilience city.

Fast Spectral Clustering for Unsupervised Hyperspectral Image Classification

Hyperspectral image classification is a challenging and significant domain in the field of remote sensing with numerous applications in agriculture, environmental science, mineralogy, and surveillance. In the past years, a growing number of advanced hyperspectral remote sensing image classification techniques based on manifold learning, sparse representation and deep learning have been proposed and reported a good performance in accuracy and efficiency on state-of-the-art public datasets. However, most existing methods still face challenges in dealing with large-scale hyperspectral image datasets due to their high computational complexity. In this work, we propose an improved spectral clustering method for large-scale hyperspectral image classification without any prior information. The proposed algorithm introduces two efficient approximation techniques based on Nyström extension and anchor-based graph to construct the affinity matrix. We also propose an effective solution to solve the eigenvalue decomposition problem by multiplicative update optimization. Experiments on both the synthetic datasets and the hyperspectral image datasets were conducted to demonstrate the efficiency and effectiveness of the proposed algorithm.

Plant disease identification from individual lesions and spots using deep learning

Deep learning is quickly becoming the standard technique for image classification. The main problem facing the automatic identification of plant diseases using this strategy is the lack of image databases capable of representing the wide variety of conditions and symptom characteristics found in practice. Data augmentation techniques decrease the impact of this problem, but those cannot reproduce most of the practical diversity. This paper explores the use of individual lesions and spots for the task, rather than considering the entire leaf. Since each region has its own characteristics, the variability of the data is increased without the need for additional images. This also allows the identification of multiple diseases affecting the same leaf. On the other hand, suitable symptom segmentation still needs to be done manually, preventing full automation. The accuracies obtained using this approach were, in average, 12% higher than those achieved using the original images. Additionally, no crop had accuracies below 75%, even when as many as 10 diseases were considered. Although the database does not cover the entire range of practical possibilities, these results indicate that, as long as enough data is available, deep learning techniques are effective for plant disease detection and recognition.

Can ensemble techniques improve coral reef habitat classification accuracy using multispectral data?

Remote sensing has potential in studies of the benthic habitat and extracting the reflectance from the data of multispectral sensors, but traditional image classification techniques cannot provide coral habitat maps with adequate accuracy. This study tested five traditional and three ensemble classification techniques on QuickBird for mapping the benthic composition of coral reefs on the Lang Tengah Island (Malaysia). The common techniques, minimum distance, maximum likelihood, K-nearest neighbour, Fisher and parallelepiped techniques were compared with ensemble classifiers, such as majority voting (MV), simple averaging, and mode combination. The per-class accuracy of the habitat detection improved in the ensemble classifiers; in particular, the MV classifier achieved 95%, 65%, 75% and 95% accuracies for coral, sparse coral, coral rubble and sand, respectively. Ensembles increased the accuracy of the habitat mapping classification by 28%, relative to conventional techniques. Thus, the ensemble techniques can be preferred over the traditional for benthic habitat mapping.

Hyperspectral image unsupervised classification by robust manifold matrix factorization

Hyperspectral remote sensing image unsupervised classification, which assigns each pixel of the image into a certain land-cover class without any training samples, plays an important role in the hyperspectral image processing but still leaves huge challenges due to the complicated and high-dimensional data observation. Although many advanced hyperspectral remote sensing image classification techniques based on supervised and semi-supervised learning had been proposed and confirmed effective in recent years, they require a certain number of high quality training samples to learn a classifier, and thus can’t work in the unsupervised manner. In this work, we propose a hyperspectral image unsupervised classification framework based on robust manifold matrix factorization and its out-of-sample extension. In order to address the high feature dimensionality of the hyperspectral image, we propose a unified low-rank matrix factorization to jointly perform the dimensionality reduction and data clustering, by which the clustering result can be exactly reproduced, which is significantly superior to the existing data clustering algorithms such as the k-means and spectral clustering. In particular, in the proposed matrix factorization, the ℓ2,1-norm is used to measure the reconstruction loss, which helps to reduce the errors brought by the possible noisy observation. The widely considered manifold regularization is also adopted to further promote the proposed model. Furthermore, we have designed a novel Augmented Lagrangian Method (ALM) based procedure to seek the local optimal solution of the proposed optimization and suggested an additional out-of-sample extension trick to make the method can deal with the large-scale hyperspectral remote sensing images. Several experimental results on the standard hyperspectral images show that the proposed method presents competitive clustering accuracy and comparative running time compared to the existing data clustering algorithms.

A new neuro-fuzzy-based classification approach for hyperspectral remote sensing images

Hyperspectral images are widely used in many applications. However, finding the appropriate hyperspectral image classification technique is a challenge. In this paper, we propose a new method by using an artificial intelligence-based method for hyperspectral image classification. The system has two parts: first, a pre-processing step, which helps the training phase to work faster; and second, the training part, which consists of calculating the neuro-fuzzy parameters. The prepared system is then applied to the classification of images. Three well-known hyperspectral datasets, including Pavia University from reflective optics system imaging spectrometer, the Botswana image from Hyperion and the Indian Pine image from airborne visible/infrared imaging spectrometer, were chosen to test the method. The final results of the experiments show that this system outperforms two classical methods of hyperspectral classification: support vector machine and spectral angle mapper. The comparison of the final results was made using two different metrics: overall accuracy and total disagreement. The proposed method increases the overall accuracy by about 5% for the Pavia University dataset, 2% for the Botswana dataset and 7% for the Indian Pine dataset. The total disagreement was reduced by about 0.01 for the Pavia University, 0.03 for the Botswana and 0.1 for the Indian Pine dataset when the proposed method was applied.

Machine learning techniques for state recognition and auto-tuning in quantum dots

Recent progress in building large-scale quantum devices for exploring quantum computing and simulation has relied upon effective tools for achieving and maintaining good experimental parameters, i.e., tuning up devices. In many cases, including quantum dot-based architectures, the parameter space grows substantially with the number of qubits, and may become a limit to scalability. Fortunately, machine learning techniques for pattern recognition and image classification, using so-called deep neural networks, have shown surprising successes for computer-aided understanding of complex systems. We propose a new paradigm for fully automated experimental initialization through a closed-loop system relying on machine learning and optimization techniques. We use deep convolutional neural networks to characterize states and charge configurations of semiconductor quantum dot arrays when only measurements of a current−voltage characteristic of transport are available. For simplicity, we model a semiconductor nanowire connected to leads and capacitively coupled to depletion gates using the Thomas−Fermi approximation and Coulomb blockade physics. We then generate labeled training data for the neural networks, and find at least 90 % accuracy for charge and state identification for single and double dots. Using these characterization networks, we can then optimize the parameter space to achieve a desired configuration of the array, a technique we call “auto-tuning”. Finally, we show how such techniques can be implemented in an experimental setting by applying our approach to an experimental dataset, and outline further problems in this domain, from using charge sensing data to extensions to full one- and two-dimensional arrays, that can be tackled with machine learning.

A survey of recent work on fine-grained image classification techniques

Image classification is a hot topic in image processing. Image classification aims to automatically classify large numbers of images. Many methods have been proposed for solving this task. Traditional methods usually leverage low-level features. Clustering is the most commonly used method of image classification. In recent years, convolutional neural networks (CNNs) is widely used in extracting deep features. Many network architectures are proposed for image classification, such as ResNeXt, Cifar10. These deep learning methods aims at fusing features of texture, color and segmentation. In this paper, we discuss the different methods and techniques of image classification, and made a detailed summary of their performance. We believe that our work plays an important role in the field of image classification.

A New Enhanced Learning Approach to Automatic Image Classification Based on Salp Swarm Algorithm

In this paper we propose a new image classification technique. According to this note that most research focuses on extraction of features in the frequency domain, location, and reduction of feature dimensions, in this research we focused on learning step in image classification. The main aim is to use the heuristic methods to increase the function of the estimator of the learning algorithm and continue to achieve the desired state, as well as categorization without user interference and automatically performed by the model produced from the above steps. So, in this paper, a new learning approach based on the Salp Swarm Algorithm was proposed that was implemented and evaluated on learning algorithm Decision Tree, K-Nearest Neighbors and Naïve Bayes. The results demonstrate the improvement of the performance of learning algorithms in all the achieved criteria by using the SSA algorithm in comparison with traditional learning algorithms. In the accuracy, sensitivity, classification error and F1 criterion, the best performance of the proposed model is using the Decision Tree learning method with values of 99.17%, 100%, 0.83% and 95.65% respectively. In the specificity and precision criterion, the best performance of the proposed model is based on K-Nearest Neighbors learning method with values of 100%.

A Comparative Study on Tumour Classification

Cancer detection is the most significant method to identify the early tumor. Enlargement of the tumor is being a huge task due to the complex characteristics of the medical images which provides high divergent, intensive and uncertain boundaries. Designing and developing computer-aided image processing systems are to help doctors improve their diagnosis and then received huge benefits over the past years. Classification is an important task within the field of computer vision. Image classification refers to the labelling of images into one of a number of predefined categories that includes image sensors, image pre-processing, object detection, object segmentation, feature extraction and object classification. Many classification techniques have been developed for image classification. The aim of literature survey is to provide a brief summary about some of common most image classification technique and comparison among them. In this survey various classification techniques are considered; Artificial Neural Network (ANN), Decision Tree (DT), Support Vector Machine (SVM), Fuzzy Classification and more.

A Literature Review on Image Processing and Classification Techniques for Agriculture Produce and Modeling of Quality Assessment system for Soybean industry Sample

A Literature Review on Image Processing and Classification Techniques for Agriculture Produce and Modeling of Quality Assessment system for Soybean industry Sample

A Proposed Arabic Text and Text Image Classification Technique Using a URL Address

Sentiment analysis is examined within natural language processing. It is instrumental in finding the sentiment (feeling) or opinion (idea) hidden within a text. This research focuses on finding sentiments in a “text image” and then classifying them whether they are desirable or not. These phrases and words refer to the perspectives of people about anything they think about it, such as services, products, governments, and social media events. In this study, the optical character recognition (OCR) algorithm was used, which is considered as a classification procedure of visual patterns that appear in the form of a digital image. Moreover, the Naïve Bayes machine learning algorithm was employed to classify these texts. These two algorithms form a hybrid system that supports our needs, especially in this day of technological advances and frequent use of websites and sharing of text images through the internet. Finally, the new vision in this work involves dealing with Arabic language texts that are transformed into images, which are extracted from a URL address and then classified into desirable and undesirable content.

CBIR system using Capsule Networks and 3D CNN for Alzheimer's disease diagnosis

Abstract Alzheimer's disease (AD) is an irreversible disorder of the brain related to loss of memory, commonly seen in the elderly and aging population. Implementation of revolutionary computer aided diagnosis techniques with Content Based Image Retrieval (CBIR) has created new potentials in Magnetic resonance imaging (MRI) in relevant image retrieval and training for detection of progression of AD in early stages. This paper proposed a CBIR system using 3D Capsule Network, 3D-Convolutional Neural Network and pre-trained 3D-autoencoder technology for early detection of Alzheimer's. A 3D-Capsule Networks (CapsNets) is capable of fast learning, even for small datasets and can effectively handle robust image rotations and transitions. It was observed that, an ensemble method using 3D-CapsNets and convolution neural network (CNN) with 3D-autoencoder, increased the detection performances comparing to Deep-CNN alone. CBIR using the proposed model was found to be up to 98.42% accurate in AD classification. Moreover, we argue that CapsNet seems to be a promising new technique for image classification, and further experiments using more robust computation resources and refined CapsNet architectures may produce better outcomes.

Implementation of Image Processing Technique for Diagnosis of Diseases

Recently, image processing techniques are widely used in several medical areas for image improvement in earlier detection and treatment stages of diseases. Medical informatics is the study that combines two medical data sources: biomedical record and imaging data. Medical image data is formed by pixels that correspond to a part of a physical object and produced by imaging modalities. discovery of medical image data methods is a challenge in the sense of getting their insight value, analyzing and diagnosing of a specific disease. Image classification plays an important role in computer-aided-diagnosis of diseases and is a big challenge on image analysis tasks. This challenge related to the usage of methods and techniques in exploiting image processing result, pattern recognition result and classification methods and subsequently validating the image classification result into medical expert knowledge. The main objective of medical images classification is to reach high accuracy to identify the name of disease. It showed the improvement of image classification techniques such as to increase accuracy and sensitivity value and to be feasible employed for computer-aided-diagnosis are a big challenge and an open research.

SEMI-AUTOMATIC ROAD NETWORK EXTRACTION FROM DIGITAL IMAGES USING OBJECT-BASED CLASSIFICATION AND MORPHOLOGICAL OPERATORS

Abstract: The demand for geospatial data concerning road network is constant, due to the wide variety of application which needs this type of data. It stands out the importance of this data in cartography update cycles, that can be obtained using automated processes of feature extraction in digital images, which are more accurate, fast and less costly than the traditional methods. In this sense, this work aimed the road network extraction from RapidEye satellite imagery, by developing a hybrid methodology using techniques of object-based image classification and morphological operators. The methodology was tested in three different sites, with images acquired in distinct dates, and the extraction process was evaluated through metrics obtained from the linear matching procedure. By the proposed extraction process, were achieved in terms of correctness and completeness the values of 92.23% and 85.15% for test site 1, the values of 79.16% and 81.06% for test site 2, and the values of 82.05% and 92.22% for test site 3, respectively. The results shown that the proposed methodology presented a good performance for semi-automatic road network extraction from Rapideye images, representing an alternative to auxiliary road network database acquisition and updating.