Explore the vast range of titles available.

This work is oriented toward video tracking of vehicles in a typical traffic environment, based on particle filters. The proposed tracking algorithm is based on simultaneous usage of three different image features - color, edge orientation, and texture. All three features are related to the contents of a rectangular window that includes both the vehicle that is tracked and local background and they are represented in the form of appropriate histograms. Based on individual estimates produced by every single feature, the resultant estimate is made by their weighted averaged. Weighting factors are adaptively changing depending on the quality of a particular feature, estimated by calculations of average similarities between the reference window and the set of windows on particles' positions. The tracking accuracies of single-feature and three-features-based filters have been verified using the set of traffic sequences illustrating the presence of typical disturbances (shadows, partial and full occlusions, maneuvering etc.).

Data hiding technologies aim to hide the existence of secret information within digital covers such as images by causing unnoticeable degradation to their quality. Reducing the image distortion and increasing the embedding capacity are the main points that the data hiding techniques revolved around. This article proposes two high payload embedding methods with high stego image quality using the Hue‐Saturation‐Value (HSV) colour model. The first method is hue‐based embedding (HBE) that employs the H plane for hiding one or two bits in non‐grey pixels. The second method uses the three HSV components, so it is called three‐planes embedding (TPE). In TPE, one bit is hidden in the least significant bit (LSB) of V of the grey pixels, one or two bits in H of the pixels having low saturation or low brightness and one bit in the LSB of S otherwise. The experiments were conducted on 25 images and the results show that HBE hides more data on average than TPE with its quality reaching 60 dB. TPE achieves quality up to 61 dB and capacity reaches 364 Kb. TPE scores the highest capacity among six state‐of‐the‐art techniques in Red‐Green‐Blue, HSV, Hue‐Saturation‐Intensity and YCbCr spaces with the highest average peak signal to noise ratio midst five of them. By embedding 60, 90, and 120 Kb, this TPE attains the best average quality amid all the methods.

For the feature extraction of red–blue–green (RGB) colour images, researchers usually deal with R, G and B channels separately to obtain three feature vectors, and then combine them together to obtain a long real feature vector. This approach does not exploit the relationships between the three channels of the colour images. Recently, attention has been paid to quaternion features, which take the relationships between channels into consideration and seem to be more suitable for representing colour images. However, there are only a few quaternion classifiers for dealing with quaternion features. To meet this requirement, a new quaternion classifier, namely, the quaternion softmax classifier is proposed, which is an extended version of the conventional softmax classifier generally defined in the complex (or real) domain. The proposed quaternion softmax classifier is applied to two of the most common quaternion features, that is, the quaternion principal components analysis feature and the colour image pixel feature. The experimental results show that the proposed method performs better than the quaternion back propagation neural network in terms of accuracy and convergence rate.

Achieving good recognition results for License plates is challenging due to multiple adverse factors. For instance, in Malaysia, where private vehicle (e.g., cars) have numbers with dark background, while public vehicle (taxis/cabs) have numbers with white background. To reduce the complexity of the problem, we propose to classify the above two types of images such that one can choose an appropriate method to achieve better results. Therefore, in this work, we explore the combination of Convolutional Neural Networks (CNN) and Recurrent Neural Networks namely, BLSTM (Bi-Directional Long Short Term Memory), for recognition. The CNN has been used for feature extraction as it has high discriminative ability, at the same time, BLSTM has the ability to extract context information based on the past information. For classification, we propose Dense Cluster based Voting (DCV), which separates foreground and background for successful classification of private and public. Experimental results on live data given by MIMOS, which is funded by Malaysian Government and the standard dataset UCSD show that the proposed classification outperforms the existing methods. In addition, the recognition results show that the recognition performance improves significantly after classification compared to before classification.

Image forensic datasets need to accommodate a complex diversity of systematic noise and intrinsic image artefacts to prevent any overfitting of learning methods to a small set of camera types or manipulation techniques. Such artefacts are created during the image acquisition as well as the manipulating process itself (e.g., noise due to sensors, interpolation artefacts, etc.). Here, the authors introduce three datasets. First, we identified the majority of camera models on the market. Then, we collected a dataset of 35,000 real images captured by these cameras. We also created the same number of digitally manipulated images. Additionally, we also collected a dataset of 2,000 ‘real‐life’ (uncontrolled) manipulated images. They are made by unknown people and downloaded from the Internet. The real versions of these images are also provided. We also manually created binary masks localising the exact manipulated areas of these images. Moreover, we captured a set of 2,759 real images formed by 32 unique cameras (19 different camera models) in a controlled way by ourselves. Here, the processing history of all images is guaranteed. This set includes categorised images of uniform areas as well as natural images that can be used effectively for analysis of the sensor noise.

Classifying pixels according to color, and segmenting the respective areas, are necessary steps in any computer vision task that involves color images. The gap between human color perception, linguistic color terminology, and digital representation are the main challenges for developing methods that properly classify pixels based on color. To address these challenges, we propose a novel method combining geometric analysis, color theory, fuzzy color theory, and multi-label systems for the automatic classification of pixels into 12 conventional color categories, and the subsequent accurate description of each of the detected colors. This method presents a robust, unsupervised, and unbiased strategy for color naming, based on statistics and color theory. The proposed model, “ABANICCO” (AB ANgular Illustrative Classification of COlor), was evaluated through different experiments: its color detection, classification, and naming performance were assessed against the standardized ISCC–NBS color system; its usefulness for image segmentation was tested against state-of-the-art methods. This empirical evaluation provided evidence of ABANICCO’s accuracy in color analysis, showing how our proposed model offers a standardized, reliable, and understandable alternative for color naming that is recognizable by both humans and machines. Hence, ABANICCO can serve as a foundation for successfully addressing a myriad of challenges in various areas of computer vision, such as region characterization, histopathology analysis, fire detection, product quality prediction, object description, and hyperspectral imaging.

Preserving hue is an important issue for colour image enhancement. Here, a hue-preserving gamut mapping method with high saturation is proposed. Experimental results with the Macbeth colour chart and natural images show vivid colour with higher subjective image quality.

Skeleton‐based action recognition algorithms have been widely applied to human action recognition. Graph convolutional networks (GCNs) generalize convolutional neural networks (CNNs) to non‐Euclidean graphs and achieve significant performance in skeleton‐based action recognition. However, existing GCN‐based models have several issues, such as the topology of the graph is defined based on the natural skeleton of the human body, which is fixed during training, and it may not be applied to different layers of the GCN model and diverse datasets. Besides, the higher‐order information of the joint data, for example, skeleton and dynamic information is not fully utilised. This work proposes a novel multi‐stream adaptive spatial‐temporal attention GCN model that overcomes the aforementioned issues. The method designs a learnable topology graph to adaptively adjust the connection relationship and strength, which is updated with training along with other network parameters. Simultaneously, the adaptive connection parameters are utilised to optimise the connection of the natural skeleton graph and the adaptive topology graph. The spatial‐temporal attention module is embedded in each graph convolution layer to ensure that the network focuses on the more critical joints and frames. A multi‐stream framework is built to integrate multiple inputs, which further improves the performance of the network. The final network achieves state‐of‐the‐art performance on both the NTU‐RGBD and Kinetics‐Skeleton action recognition datasets. The simulation results prove that the proposed method reveals better results than existing methods in all perspectives and that shows the superiority of the proposed method.

The object contour is a significant cue for identifying and categorising objects. The current work is motivated by indicative researches that attribute object contours to edge information. The spatial continuity exhibited by the edge pixels belonging to the object contour make these different from the noisy edge pixels belonging to the background clutter. In this study, the authors seek to quantify the object contour from a relative count of the adjacent edge pixels that are oriented in the four possible directions, and measure using exponential functions the continuity of each edge over the next adjacent pixel in that direction. The resulting computationally simple, low-dimensional feature set, called as ‘edge continuity features’, can successfully distinguish between object contours and at the same time discriminate intra-class contour variations, as proved by the high accuracies of object recognition achieved on a challenging subset of the Caltech-256 dataset. Grey-to-RGB template matching with City-block distance is implemented that makes the object recognition pipeline independent of the actual colour of the object, but at the same time incorporates colour edge information for discrimination. Comparison with the state-of-the-art validates the efficiency of the proposed approach.

Surgical instrument detection in robot-assisted surgery videos is an import vision component for these systems. Most of the current deep learning methods focus on single-tool detection and suffer from low detection speed. To address this, the authors propose a novel frame-by-frame detection method using a cascading convolutional neural network (CNN) which consists of two different CNNs for real-time multi-tool detection. An hourglass network and a modified visual geometry group (VGG) network are applied to jointly predict the localisation. The former CNN outputs detection heatmaps representing the location of tool tip areas, and the latter performs bounding-box regression for tool tip areas on these heatmaps stacked with input RGB image frames. The authors’ method is tested on the publicly available EndoVis Challenge dataset and the ATLAS Dione dataset. The experimental results show that their method achieves better performance than mainstream detection methods in terms of detection accuracy and speed.