International Journal of Pattern Recognition and Artificial Intelligence

Tactile sensing has recently been used in pattern recognition technology for pilots’ posture information and environmental information. Human tactile sensing is limited, however, through control of the spatial distribution and vibration intensity of each contact in the tactile stimulation array, the accuracy, convenience, and comfort of the tactile device can be comprehensively improved. Moreover, the recognition rate of most current flight posture information methods is low. In this paper, the principle of vibration haptic coding is optimized. A combined coding scheme of “vibration $\mathrm{\text{position}}+\mathrm{\text{vibration}}\mathrm{\text{mode}}+\mathrm{\text{vibration}}$ sequence” is used to recognize pilot’s flight posture. A novel triangular coding scheme is proposed for the first time. Compared to other commonly used coding schemes such as “needle scheme“ and “rectangle scheme”, experimental results show that the triangular coding scheme is 1.5% more accuracy with response time reduced by nearly 1.25s in recognition pilots’ flight posture information.

A practically deployable gesture recognition system is developed using a robust hand detection method implemented using a motion-based image segmentation process and a two-level bare hand classification model, which is integrated with a gesture classification system of 58 gestures using new robust features. Since detection of bare hand is affected by nonideal conditions, multiple color-texture features are analyzed in this study. In the second stage of the system, 18 new ASCII characters are introduced and analyzed along with the existing 40 characters (alphabets, numbers, and arithmetic operators). New 15 dimensional features are introduced along with the existing features to enhance the classification accuracy of the gestures. Significance of features statistically tested using one-way analysis of variance (ANOVA), Kruskal–Wallis and Friedman test, which are sequentially ranked and evaluated using incremental feature selection (IFS) method. Performance of the proposed hand detection system is observed to be 12.5% higher than the existing hand detection system under clean conditions, while 46.4% higher under the nonideal conditions. Performance of 58 gestures classification model has improved by 12.08% (Naïve Bayes), 8.86% (ELM), 10.83% (SVM), 8.02% ($k$NN), and 6.61% (ANN) after using the new features. Majority voting-based classifier fusion method further improves the performance of the gesture recognition system by 3.88%, which is validated by Turkey’s HSD test.

Due to the deforming and dynamically changing textile fibers, the quality assurance of cleaned industrial textiles is still a mostly manual task. Usually, textiles need to be spread flat, in order to detect defects using computer vision inspection methods. Already known methods for detecting defects on such inhomogeneous, voluminous surfaces use mainly supervised methods based on deep neural networks and require lots of labeled training data. In contrast, we present a novel unsupervised method, based on SURF keypoints, that does not require any training data. We propose using their location, number and orientation in order to group them into geographically close clusters. Keypoint clusters also indicate the exact position of the defect at the same time. We furthermore compared our approach to supervised methods using deep learning. The presented processing pipeline shows how normalization and classification methods need to be combined, in order to reliably detect fiber defects such as cuts and holes. We evaluate the performance of our system in real-world settings with images of piles of textiles, taken in stereo vision. Our results show that our novel unsupervised classification method using keypoint clustering achieves comparable results to other supervised methods.

In this paper, we propose a fully autonomous density-based clustering algorithm named ‘Ocean’, which is inspired by the oceanic landscape and phenomena that occur in it. Ocean is an improvement over conventional algorithms regarding both distance metric and the clustering mechanism. Ocean defines the distance between two categories as the difference in the relative densities of categories. Unlike existing approaches, Ocean neither assigns the same distance to all pairs of categories, nor assigns arbitrary weights to matches and mismatches between categories that can lead to clustering errors. Ocean uses density ratios of adjacent regions in multidimensional space to detect the edges of the clusters. Ocean is robust against clusters of identical patterns. Unlike conventional approaches, Ocean neither makes any assumption regarding the data distribution within clusters, nor requires tuning of free parameters. Empirical evaluations demonstrate improved performance of Ocean over existing approaches.

Cross-modal retrieval aims to retrieve the related items from various modalities with respect to a query from any type. The key challenge of cross-modal retrieval is to learn more discriminative representations between different category, as well as expand to an unseen class retrieval in the open world retrieval task. To tackle the above problem, in this paper, we propose a prototype learning-based discriminative feature learning (PLDFL) to learn more discriminative representations in a common space. First, we utilize a prototype learning algorithm to cluster these samples labeled with the same semantic class, by jointly taking into consideration the intra-class compactness and inter-class sparsity without discriminative treatments. Second, we use the weight-sharing strategy to model the correlations of cross-modal samples to narrow down the modality gap. Finally, we apply the prototype to achieve class-incremental learning to prove the robustness of our proposed approach. According to our experimental results, significant retrieval performance in terms of mAP can be achieved on average compared to several state-of-the-art approaches.

Recent research has shown that features obtained from pretrained Convolutional Neural Network (CNN) models can be promptly applied to a variety of problems they were not originally designed to solve. This concept, often referred to as Transfer Learning (TL), is a common practice when labeled data is limited. In some fields, such as video anomaly detection, TL is still an underexplored subject in the sense that it is not clear whether the architecture of the pretrained CNN model impacts on the video anomaly detection performance. In order to clarify this issue, we perform an extensive benchmark using 12 different pretrained CNN models on ImageNet as feature extractors and apply the features obtained to seven video anomaly detection benchmark datasets. This work presents some interesting findings about video anomaly detection using TL. The highlights of our findings were revealed by our experiments, which have shown that a simple classification process using One-Class Support Vector Machines yields similar results to state-of-the-art models. Moreover, a statistical analysis suggests that architectural differences are negligible when choosing a pretrained model for video anomaly detection, since all models presented similar performance. At last, we present an in-depth visual analysis of the Avenue dataset, and reveal several aspects that may be limiting the performance of state-of-the-art video anomaly detection methods.

Watermelon is a crop susceptible to diseases. Rapid and effective detection of watermelon diseases is of great significance to ensure the yield of watermelon. Aiming at the interference of the environment and obstacles in the natural environment, resulting in low target detection accuracy and poor robustness, this paper takes watermelon leaves as the research object, considering anthracnose, leaf blight, leaf spot and normal leaves as examples. A disease recognition method based on deep learning is proposed. This paper has improved the pre-selected box setting formula of the SSD model and tested it in multiple SSD models. Experiments show that the average accuracy of the final SSD768 model is 92.4%, and the average accuracy of the IOU is 88.9%. It shows that this method can be used to detect watermelon diseases in natural environment.

Infrared target tracking technology is one of the core technologies in infrared imaging guidance systems and is also a hot research topic. The problem of particle degradation could be always found in traditional particle filtering, and a large number of particles are additionally required for accurate estimation. It is difficult to meet the requirements of a modern infrared imaging guidance system for accurate target tracking. To solve the problem of particle degradation and improve the performance of infrared target tracking, the extended Kalman filter and genetic algorithm are introduced into particle filtering, and an improved algorithm for infrared target tracking is proposed in this paper. In the framework of a particle filter algorithm, the Gaussian distribution for each particle is generated and propagated by a separate extended Kalman filter to improve the sampling accuracy and effectiveness of the probability density function of particles. Genetic algorithm is used to perform a resampling process to solve particle degradation and ensure the diversity of particle states in particle swarm. Simulation results show that the improved tracking algorithm based on improved particle filtering proposed in this paper can effectively solve the phenomenon of particle degradation and track the infrared target.

Terrain synthesis has been a hot topic in the field of computer graphics and image processing. However, there are still issues in terrain synthesis where synthesis results are difficult to control and not realistic enough. To address these problems, this paper proposes an interactive terrain elevation map generation method based on the synthesis of a single sample terrain elevation map. First, we propose a method to extract the skeleton from a terrain elevation map and a user sketch. Second, we construct a skeleton sample feature map based on the terrain elevation map and the user sketch. Finally, we propose a matching cost function to match image patches of the terrain sample and the user sketch. The proposed method can obtain a synthesis result containing the features of both the terrain sample and the user sketch, and then generates a synthetic terrain elevation map. The experimental results demonstrate the effectiveness of the proposed method, where the synthesized results can meet the needs of users.

In order to enhance the sense of reality haptic display based on image, it is widely expected to express various characteristics of the objects in the image using different kinds of haptic feedback. To this end, a multi-mode haptic display method of image was proposed in this paper, including the multi-feature extraction of image and the image expression with various types of haptic rendering. First, the device structure integrating force and vibrotactile feedbacks was designed for multi-mode haptic display. Meanwhile, the three-dimensional geometric shape, detail texture and outline of the object in the image were extracted by various image processing algorithms. Then, a rendering method for the object in the image was proposed based on the psychophysical experiments on the piezoelectric ceramic actuator. The 3D geometric shape, detail texture and outline of the object were rendered by force and vibration tactile feedbacks, respectively. Finally, these three features of the image were haptic expressed simultaneously by the integrated device. Haptic perception experiment results show that the multi-mode haptic display method can effectively improve the authenticity of haptic perception.

In order to detect the type of vehicle seat and the missing part of the spring hook, this paper proposes an improved RANSAC-SURF method. First, the image is filtered by a Gauss filter. Second, an improved RANSAC-SURF algorithm is used to detect the types of vehicle seats. Extract the feature points of vehicle seats. The feature points are matched according to the improved RANSAC-SURF algorithm. Third, the image distortion of the vehicle seat is corrected by the method of perspective transformation. Determine whether the seat’s spring hook is missing or not according to the absolute value of the gray difference between the image collected by the camera and the image of the normal installation. The experimental results show that the MSE of the Gauss filter under a 5 $*$ 5 template is 19.0753, and the PSNR is 35.3261, which is better than that of the mean filter and the median filter. The total matching logarithm of feature points and the number of intersection points are 188 and 18, respectively, in the improved RANSAC-SURF matching algorithm.

With the development of deep learning technologies, object detection algorithms have made significant progress in terms of detection speed and detection performance. However, the detection speed of current detection networks still does not meet the requirements of real-world applications in some scenarios. In this paper, we propose a faster non-maximum suppression (FNMS) algorithm that reduces the processing time by a large margin while achieving the same detection precision compared with the traditional non-maximum suppression (NMS) algorithm. Moreover, an attempt is made to adopt additional lightweight network structures to improve the speed of the detection network. By combining our FNMS algorithm with other network optimization strategies, we are able to improve the detection speed of YOLO v3 on the DOTA dataset by 165%.

Deep learning technology has greatly improved the performance of target tracking, but most recently developed tracking algorithms are short-term tracking algorithms, which cannot meet the actual engineering needs. Based on the Siamese network structure, this paper proposes a long-term tracking framework with a persistent tracking capability. The global proposal module extends the search area globally through the construction of a feature pyramid. The local regression module is mainly responsible for the confidence evaluation of the candidate regions and for performing more accurate bounding box regression. To improve the discriminative ability of the regression network, the error samples are eliminated by synthesizing the temporal information and are then classified through a verification module in advance. Experiments on the VOT long-term tracking dataset and the UAV20L aerial dataset show that the proposed algorithm achieves state-of-the-art performance.

Chest X-ray images are notoriously difficult to analyze due to the noisy nature. Automatic identification of pneumonia on medical images has attracted intensive study recently. In this paper, a novel joint-task architecture that can learn pneumonia classification and segmentation simultaneously is presented. Two modules, including an image preprocessing module and an attention module, are developed to improve both the classification and segmentation accuracies. Results from the experiments performed on the massive dataset of the Radiology Society of North America have confirmed its superiority over the other existing methods. The classification test accuracy is improved from 0.89 to 0.95, and the segmentation model achieves an improved mean precision result of 0.58–0.78. Finally, two weakly supervised learning methods, class-saliency map and Grad-CAM, are used to highlight the corresponding pixels or areas which have significant influence on the classification model, such that the refined segmentation can focus on the correct areas with high confidence.

White blood cells (WBCs) play a main role in identifying the health condition and disease characteristics of a normal person. An automated classification system is capable of recognizing white blood cells that may help doctors to diagnose several diseases like malaria, anemia, leukemia, etc. Automated blood cell analysis allows fast and accurate outcomes and often involves broad data without performance negotiation. The state-of-the-art systems use a lot of different stages (feature extraction, segmentation, pre-processing, etc.) to provide the automated blood cell analysis using blood smear images which is a lengthy process. To overcome these problems, this paper presents an efficient peripheral blood cell image recognition and classification using a combination of the salp swarm algorithm and the cat swarm optimization (SSPSO) algorithm-based optimized convolutional neural networks (SSPSO-CNN) method. This paper uses the CNN approach to classify five peripheral blood cells such as eosinophil, basophil, lymphocytes, monocytes, and neutrophils without any human intervention. The other objective of this paper is to propose an improved version of salp swarm optimizer (SSO) using particle swarm optimization (PSO) to attain competitive classification performance over the database of the blood cell images. In this paper, the CNN uses VGG19 architecture for training purposes. The accuracy of the classification achieved with VGG19 models is 98%. The proposed model based on the CNN approach optimized by SSPSO achieves high classification accuracy and provides automatic peripheral blood cell classification. This method establishes the fine-tuning process to develop a classifier trained using 10 674 images obtained from medical practice. The proposed method augmented the performance in terms of high precision and $F$1-score and obtained an overall classification accuracy of 99%.

Routing algorithm is an important factor and a key technology for Flying Ad hoc Networks (FANETs), which can safeguard FANET network communication. In FANETs, the fast-changing nature of FANET network topology makes the traditional Mobile Ad-hoc Networks (MANET) network routing algorithm not directly usable, which leads to challenges for the design of routing algorithms. In this paper, when the FANET nodes need to communicate, the route search program performs route search based on the ant colony algorithm, so as to obtain a stable route with high efficiency. Making use of NS3 for simulation of ACA, Dynamic Source Routing (DSR), and Ad hoc On demand Distance Vector Routing (AODV), the results of simulation show that ACA can improve FANET performance.