On-line Classification System Research Articles

Spectral imaging and chemometrics applied at phenotyping in seed science studies: a systematic review

AbstractThe evaluation of the genetic quality of a seed lot is crucial for the quality control process in its production and commercialization, as well as in the identification of superior genotypes and the verification of the correct crossing in plant breeding programmes. Current techniques, based on the identification of seed morphological characteristics, require skilled analysts, while biochemical methods are time-consuming and costly. The application of spectral imaging analysis, which combines digital imaging with spectroscopy, is gaining ground as a fast, accurate and non-destructive method. The success of this technique is closely linked to chemometric techniques, which use statistical and mathematical tools in data processing. The aim of the work was to evaluate the main procedures in terms of spectral image analysis and chemometric procedures applied in seed phenotyping and its practical application. A systematic review was conducted using the PRISMA methodology, in which a total of 1304 articles were identified and screened to the inclusion of 44 articles pertaining to the scope. It was concluded that spectral image analysis has a high ability to classify seeds of different genotypes (93.33%) in a range of situations: between cultivars; hybrids and progenitors; and hybrids and lines, as well as in the separation of coated seeds. Accurate classification can be obtained by different strategies, such as the choice of the equipment type, the spectrum range and extra features, guided by the characteristics of the species, as well as in the choice of algorithms and dimensionality reduction procedures for the optimization of models when there is a large amount of data. Despite the fact that the practical application of this technique in seed phenotyping still needs to be developed for use in laboratories with large volumes of analyses, lots, genotypes and harvests. Research has been accelerated to overcome the practical challenges of this method, as seen in works using model update algorithms, online classification systems, and real-time classification maps. Thus, there are strong indications that the application of multispectral image analysis will reach the routine of seed analysis laboratories.

On-line recognition of emotions via electroencephalography

Automated pattern recognition of brain signals can bring about new experiences, enhancing applications in a wide array of areas. One of its fields of study is the recognition of emotions via electroencephalography (EEG), which shows exclusive advantages compared to other methods. However, research with brain-computer interfaces (BCI) is usually structured in two sequential stages: data collection and data analysis. These stages leave a gap in the perspective of a functional system in a production environment since the practitioner needs to wait a considerable length of time until they can see the results of the current activity. An on-line classification system of emotions (positive, neutral, and negative) was developed using open resources in this work. Five machine learning models were trained with the SEED IV dataset, which is labeled with different emotions. The models were trained and tested using nested cross-validation and grid search to obtain the best hyperparameters. The algorithm implementation in Python was integrated with the OpenBCI software to capture the EEG signals, process them, and command the simulations. The best average accuracy obtained for a single subject was 76.19%, and the average accuracy for all subjects was 57.07%. The average execution time for signal processing and prediction, combined, was around one millisecond, which demonstrates the potential for applications with real-time characteristics.

3DeepM: An Ad Hoc Architecture Based on Deep Learning Methods for Multispectral Image Classification

Current predefined architectures for deep learning are computationally very heavy and use tens of millions of parameters. Thus, computational costs may be prohibitive for many experimental or technological setups. We developed an ad hoc architecture for the classification of multispectral images using deep learning techniques. The architecture, called 3DeepM, is composed of 3D filter banks especially designed for the extraction of spatial-spectral features in multichannel images. The new architecture has been tested on a sample of 12210 multispectral images of seedless table grape varieties: Autumn Royal, Crimson Seedless, Itum4, Itum5 and Itum9. 3DeepM was able to classify 100% of the images and obtained the best overall results in terms of accuracy, number of classes, number of parameters and training time compared to similar work. In addition, this paper presents a flexible and reconfigurable computer vision system designed for the acquisition of multispectral images in the range of 400 nm to 1000 nm. The vision system enabled the creation of the first dataset consisting of 12210 37-channel multispectral images (12 VIS + 25 IR) of five seedless table grape varieties that have been used to validate the 3DeepM architecture. Compared to predefined classification architectures such as AlexNet, ResNet or ad hoc architectures with a very high number of parameters, 3DeepM shows the best classification performance despite using 130-fold fewer parameters than the architecture to which it was compared. 3DeepM can be used in a multitude of applications that use multispectral images, such as remote sensing or medical diagnosis. In addition, the small number of parameters of 3DeepM make it ideal for application in online classification systems aboard autonomous robots or unmanned vehicles.

Online Application of a Hyperspectral Imaging System for the Sorting of Adulterated Almonds

Almonds are nutrient-rich nuts. Due to their high level of consumption and relatively high price, their production is targeted for illegal practices, with the intention of earning more profit. The most common adulterants are based on superficial matching, and as an adulterant, the apricot kernel is comparatively inexpensive and almost identical in color, texture, odor, and other physicochemical characteristics to almonds. In this study, a near-infrared hyperspectral imaging (NIR-HSI) system in the wavelength range of 900–1700 nm synchronized with a conveyor belt was used for the online detection of added apricot kernels in almonds. A total of 448 samples from different varieties of almonds and apricot kernels (112 × 4) were scanned while the samples moved on the conveyor belt. The spectral data were extracted from each imaged nut and used to develop a partial least square discrimination analysis (PLS-DA) model coupled with different preprocessing techniques. The PLS-DA model displayed over a 97% accuracy for the validation set. Additionally, the beta coefficient obtained from the developed model was used for pixel-based classification. An image processing algorithm was developed for the chemical mapping of almonds and apricot kernels. Consequently, the obtained model was transferred for the online sorting of seeds. The online classification system feedback had an overall accuracy of 85% for the classification of nuts. However, the model presented a relatively low accuracy when evaluated in real-time for online application, which might be due to the rough distribution of samples on the conveyor belt, high speed, delaying time in suction, and lighting variations. Nevertheless, the developed online prototype (NIR-HSI) system combined with multivariate analysis exhibits strong potential for the classification of adulterated almonds, and the results indicate that the system can be effectively used for the high-throughput screening of adulterated almond nuts in an industrial environment.

Development of online classification system for construction waste based on industrial camera and hyperspectral camera.

Construction waste is a serious problem that should be addressed to protect environment and save resources, some of which have a high recovery value. To efficiently recover construction waste, an online classification system is developed using an industrial near-infrared hyperspectral camera. This system uses the industrial camera to capture a region of interest and a hyperspectral camera to obtain the spectral information about objects corresponding to the region of interest. The spectral information is then used to build classification models based on extreme learning machine and resemblance discriminant analysis. To further improve this system, an online particle swarm optimization extreme learning machine is developed. The results indicate that if a near-infrared hyperspectral camera is used in conjunction with an industrial camera, construction waste can be efficiently classified. Therefore, extreme learning machine and resemblance discriminant analysis can be used to classify construction waste. Particle swarm optimization can be used to further enhance the proposed system.

Integrating new classes on the fly in evolving fuzzy classifier designs and their application in visual inspection

In this paper, we address the problem of integrating new classes on the fly into on-line classification systems. The main focus is on visual inspection tasks, although the concepts proposed in this paper can easily be applied to any other on-line classification systems. We use evolving fuzzy classifiers (EFCs), which can adapt their structure and update their parameters in incrementally due to embedded on-line adaptable classifier learning engines. We consider two different model architectures — classical single model and an all-pairs approach that uses class information to decompose the classification problem into several smaller sub-problems. The latter technique is essential for establishing new classes quickly and efficiently in the classifier, and for reducing class imbalance. Methodological novelties are (i) making appropriate structural changes in the EFC whenever a new class appears while operating in a single-pass incremental manner and (ii) estimating the expected change in classifier accuracy on the older classes. The estimation is based on an analysis of the impact of new classes on the established decision boundaries. This is important for operators, who are already familiar with an established classifier, the accuracy of which is known. The new concepts are evaluated in a real-world visual inspection scenario, where the main task is to classify event types which may occur dynamically on micro-fluidic chips and may reduce their quality. The results show stable performance of established classifiers and efficient (low number of samples requested) as well as fast integration (steeply rising accuracy curves) of new event types (classes).

Visual traffic surveillance framework: classification to event detection

Visual traffic surveillance using computer vision techniques can be noninvasive, automated, and cost effective. Traffic surveillance systems with the ability to detect, count, and classify vehicles can be employed in gathering traffic statistics and achieving better traffic control in intelligent transportation systems. However, vehicle classification poses a difficult problem as vehicles have high intraclass variation and relatively low interclass variation. Five different object recognition techniques are investigated: principal component analysis (PCA)+difference from vehicle space, PCA+difference in vehicle space, PCA+support vector machine, linear discriminant analysis, and constellation-based modeling applied to the problem of vehicle classification. Three of the techniques that performed well were incorporated into a unified traffic surveillance system for online classification of vehicles, which uses tracking results to improve the classification accuracy. To evaluate the accuracy of the system, 31 min of traffic video containing multilane traffic intersection was processed. It was possible to achieve classification accuracy as high as 90.49% while classifying correctly tracked vehicles into four classes: cars, SUVs/vans, pickup trucks, and buses/semis. While processing a video, our system also recorded important traffic parameters such as the appearance, speed, trajectory of a vehicle, etc. This information was later used in a search assistant tool to find interesting traffic events.

Field testing of a system for online classification of beef carcasses for longissimus tenderness using visible and near-infrared reflectance spectroscopy1,2

The present experiments were conducted to field test a system optimized for online prediction of beef LM tenderness based on visible and near-infrared (VISNIR) spectroscopy and to develop and validate a model for prediction of tenderness that would be unbiased by normal variation in bloom time before application of VISNIR. For both Exp. 1 and 2, slice shear force (SSF) was measured on fresh (never frozen) steaks at 14 d postmortem. Carcasses with VISNIR-predicted SSF ≤15 kg were classified as VISNIR predicted tender and carcasses with VISNIR-predicted SSF >15 kg were classified as VISNIR not predicted tender. In Exp. 1, spectroscopy was conducted online, during carcass grading, at 3 large-scale commercial fed-beef processing facilities. Each carcass (n = 1,155) was evaluated immediately after ribbing and again when the carcass was graded. For model development and validation, carcasses were blocked by plant and observed SSF. One-half of the carcasses (n = 579) were assigned to a calibration data set, which was used to develop regression equations, and one-half of the carcasses (n = 576) were assigned to a prediction data set, which was used to validate the regression equations. Carcasses predicted tender by VISNIR spectroscopy had smaller (P < 10(-19)) mean LM SSF values at 14 d postmortem in the calibration (13.9 vs. 16.5 kg) and prediction (13.8 vs. 16.4 kg) data sets than did carcasses not predicted tender by VISNIR spectroscopy. Relative to carcasses not predicted tender by VISNIR, a decreased percentage of carcasses predicted tender by VISNIR had LM SSF >25 kg in the calibration (2.0 vs. 7.8%) and prediction (0.8 vs. 8.0%) data sets. In Exp. 2, carcasses (n = 4,204) were evaluated with VISNIR online at 6 commercial fed-beef processing facilities on 38 production days. The carcasses predicted tender by VISNIR spectroscopy had decreased mean LM SSF values at 14 d postmortem (16.3 vs. 19.9 kg; P < 10(-87)), longer sarcomere lengths (1.77 vs. 1.72 µm; P < 10(-10)), and a greater percentage of desmin degraded (42 vs. 34%; P < 10(-5)) by 14 d postmortem. Relative to carcasses not predicted tender by VISNIR, a decreased percentage of carcasses predicted tender by VISNIR had LM SSF >25 kg (4.9 vs. 21.3%). The present experiments resulted in development and independent validation of a robust method to noninvasively predict LM tenderness of grain-fed beef carcasses. This technology could facilitate tenderness-based beef merchandising systems.

Hybrid active learning for reducing the annotation effort of operators in classification systems

Active learning is understood as any form of learning in which the learning algorithm has some control over the input samples due to a specific sample selection process based on which it builds up the model. In this paper, we propose a novel active learning strategy for data-driven classifiers, which is based on unsupervised criterion during off-line training phase, followed by a supervised certainty-based criterion during incremental on-line training. In this sense, we call the new strategy hybrid active learning. Sample selection in the first phase is conducted from scratch (i.e. no initial labels/learners are needed) based on purely unsupervised criteria obtained from clusters: samples lying near cluster centers and near the borders of clusters are expected to represent the most informative ones regarding the distribution characteristics of the classes. In the second phase, the task is to update already trained classifiers during on-line mode with the most important samples in order to dynamically guide the classifier to more predictive power. Both strategies are essential for reducing the annotation and supervision effort of operators in off-line and on-line classification systems, as operators only have to label an exquisite subset of the off-line training data resp. give feedback only on specific occasions during on-line phase. The new active learning strategy is evaluated based on real-world data sets from UCI repository and collected at on-line quality control systems. The results show that an active learning based selection of training samples (1) does not weaken the classification accuracies compared to when using all samples in the training process and (2) can out-perform classifiers which are built on randomly selected data samples.

Technical Note: A Condition Classification System for Reciprocating Compressors

This paper proposes a classification system for damage detection that combines the discrete wavelet transform, a statistical methodology, and neural networks. In the proposed system, vibration signals of compressors are acquired as initial system inputs. The condition features are extracted by statistical moments of wavelet coefficients before entering the networks. The proposed on-line classification system can automatically classify a normal or faulty compressor. The system has been implemented to increase product reliability and reduce downtime for a large production facility.

Automating the Library of Congress Classification Scheme

Potential uses for classification data in machine readable form and reasons for the development of a standard, the USMARC Format for Classification Daata, which allows for classification data to interact with other USMARC bibliographic and authority data are discussed. The development, structure, content, and use of the standard is reviewed with implementation decisions for the Library of Congress Classification scheme noted. The author examines the implementation of USMARC classification at LC, the conversion of the schedules, and the functionality of the software being used. Problems in the effort are explored, and enhancements desired for the online classification system are considered.

DECISION TREES FOR AUTOMATED IDENTIFICATION OF COSMIC-RAY HITS IN HUBBLE SPACE TELESCOPE IMAGES

ABSTRACT We have developed several algorithms for classifying objects in astronomical images. These algorithms have been used to label stars, galaxies, cosmic rays, plate defects, and other types of objects in sky surveys and other image databases. Our primary goal has been to develop techniques that classify with high accuracy, in order to ensure that celestial objects are not stored in the wrong catalogs. In addition, classification time must be fast due to the large number of classifications and to future needs for on-line classification systems. This paper reports on our results from using decision tree classifiers to identify cosmic ray hits in Hubble Space Telescope images. This method produces classifiers with over 95% accuracy using data from a single, upaired image. Our experiments indicate that this accuracy will get even higher if methods for eliminating background noise improve.