Binary Descriptors Research Articles

Vegetation Cover Estimation Using Convolutional Neural Networks

Vegetation is an important parameter in all bio- and ecosystems, and it should be monitored to conserve and restore the environment. This paper presents a design and implementation of a compact system for vegetation cover monitoring, which consists of a small unmanned aerial vehicle (UAV) equipped with four different cameras (RGB, NIR, NDVI, and red). These cameras simultaneously record videos and wirelessly send them to a ground station that applies on them a set of algorithms to construct a mosaic representing the covered area and segment vegetative regions. Mosaicing is performed using a fast multi-threaded approach based on binary descriptors. For segmenting vegetation, two scenarios are tested by using convolutional neural networks (CNNs). The first scenario trained a CNN using the images obtained from the four cameras, and then used it with the constructed mosaics. In the second scenario, we trained individual CNNs using images from each of the four cameras individually. The performance of the former scenario exceeded the others from the perspective of accuracy. Moreover, it proved to be highly comparable to previous approaches utilizing level sets, while the processing time is reduced. The proposed approach obtained high accuracy in terms of the Dice similarity coefficient (97 %), which demonstrates its favorable performance.

Unified dynamic texture segmentation system based on local and global spatiotemporal techniques

Dynamic texture (DT) is temporal extension of static texture. There are two broad types of dynamic textures: natural and manmade. The examples of dynamic textures are waving tree, sea water, fountain, traffic, moving crowd, etc. Dynamic texture segmentation is a technique used to separate the moving objects from stationary content in the video. Majority of the techniques give good results either for natural or for manmade dynamic textures. The proposed approach for DT Segmentation is combination of local spatiotemporal technique, i.e., local binary pattern-Weber local descriptor and global spatiotemporal technique, i.e., contourlet transform. The local spatiotemporal technique considers the appearance and motion of the object for segmentation. The technique is computationally less complex than optical flow and gives good results for manmade dynamic textures. The global spatiotemporal technique is based on Laplacian pyramid and directional filters. It gives good results for natural dynamic textures. The proposed technique discussed in this paper is unified approach for any type of dynamic texture. The regions of images commonly segmented by both the techniques are considered as the final segmented output. The proposed system works equally well on any kind of dynamic texture.

Unified dynamic texture segmentation system based on local and global spatiotemporal techniques

Dynamic texture (DT) is temporal extension of static texture. There are two broad types of dynamic textures: natural and manmade. The examples of dynamic textures are waving tree, sea water, fountain, traffic, moving crowd, etc. Dynamic texture segmentation is a technique used to separate the moving objects from stationary content in the video. Majority of the techniques give good results either for natural or for manmade dynamic textures. The proposed approach for DT Segmentation is combination of local spatiotemporal technique, i.e., local binary pattern-Weber local descriptor and global spatiotemporal technique, i.e., contourlet transform. The local spatiotemporal technique considers the appearance and motion of the object for segmentation. The technique is computationally less complex than optical flow and gives good results for manmade dynamic textures. The global spatiotemporal technique is based on Laplacian pyramid and directional filters. It gives good results for natural dynamic textures. The proposed technique discussed in this paper is unified approach for any type of dynamic texture. The regions of images commonly segmented by both the techniques are considered as the final segmented output. The proposed system works equally well on any kind of dynamic texture.

Fingerprint indexing via BRIEF minutia descriptors

We use BRIEF binary local image descriptors as minutia descriptors for indexing of biometric fingerprint databases. Tests with varying descriptor size and parametrisation are performed on a proprietary database. Compared with the speed of an implementation of conventional minutiae matching, we find that BRIEF descriptors are fast enough for database indexing. The tested descriptors outperform two other image descriptors (LBP, HoG) from recent literature with respect to matching rates and average penetration rates.

SPATIAL AND STATISTICAL ANALYSIS AND PROCESSING OF FRAGMENTS ENSEMBLE OF IMAGE STRUCTURAL DESCRIPTION IN RECOGNITION OF VISUAL OBJECTS

The task of visual object recognition in computer vision systems using the feature space of special image points is solved based on the statistical distributions for a system of fragments of image description. On the basis of cross-correlation processing of the distribution options for the one-bit blocks of the binary descriptors, a new system of integrated features was constructed for speed-efficient calculation of the description relevance in the process of object recognition. By software simulation, an experimental evaluation of the computation time of the relevance value in the constructed feature space is performed in comparison with the use of full distribution and the traditional voting approach for a set of descriptors. The effectiveness of the proposed system of features in terms of a significant increase in the performance speed with an adequate level of recognition quality is confirmed.

GCNv2: Efficient Correspondence Prediction for Real-Time SLAM

In this letter, we present a deep learning-based network, GCNv2, for generation of keypoints and descriptors. GCNv2 is built on our previous method, GCN, a network trained for 3D projective geometry. GCNv2 is designed with a binary descriptor vector as the ORB feature so that it can easily replace ORB in systems such as ORB-SLAM2. GCNv2 significantly improves the computational efficiency over GCN that was only able to run on desktop hardware. We show how a modified version of ORBSLAM2 using GCNv2 features runs on a Jetson TX2, an embedded low-power platform. Experimental results show that GCNv2 retains comparable accuracy as GCN and that it is robust enough to use for control of a flying drone. Source code is available at: https://github.com/jiexiong2016/GCNv2_SLAM.

Discursive Strategies of Oromara Integration: A Critical Discourse Analysis of Abiy Ahmed’s Political Rhetoric

This article analyzes the political rhetoric Abiy Ahmed used to advocate integration between the Oromo and Amhara ethnic groups, recasting it as “Oromara.” Its objectives are: identifying the rhetorical devices and discursive strategies used in the integration of the Oromara, examining how integration of the groups is discursively constructed, describing the message characteristics of the interview texts, and explaining how political rhetoric, power and ideology are linked to play advocacy role for further reinforcing the integration of the groups. To attain these objectives, the interview Abiy had with Amhara television journalist on 13 November, 2017 was selected using purposive sampling. The interview data was changed into written forms using line based transcription system, and analyzed thematically. Fairclough’s (1992) model of discourse analysis was used in the analysis. The findings show that rhetorical devices (simile, repetition, parallelism and analogy, metaphor and metonymy) are employed in the text. The macro-discursive strategies used in the interview include accommodative, normalization, constructive and genericizing. The interview text communicated both negative (hostility, revenge, detestation, distrust, conflict and retaliation) and positive (integration, trust, forgiveness, peace and adoration) messages. The negative descriptors portray unity of the groups as a threat for political system designed for the last 27 years, whereas the positive descriptors indicate that discourse of the Oromara is a key to bring systemic change in the country. The binary descriptors explain the divide and rule system constructed by the monopolized political power on the one hand, and the power of integration to topple the system on the other. Keywords: political rhetoric, rhetorical device, discursive strategy, critical discourse analysis

3D Surface Matching by a Voxel-Based Buffer-Weighted Binary Descriptor

3D surface matching by local feature descriptors is a fundamental task in 3D object registration, recognition, and retrieval. In view of 2D projected descriptors’ information compression and 3D directly voxelized descriptors’ sensitivity to density variation and boundary, this paper first proposes a voxel-based buffer-weighted binary descriptor, named VBBD. After local surfaces of detected keypoints are voxelized, each voxel’s buffer region is established, and the buffer-weighted Gaussian kernel density is calculated. If the current voxel’s buffer-weighted density is larger than its local surface’s average density, the voxel is labeled 1, otherwise, it is 0. The proposed descriptor has several merits: 1) direct acquisition of 3D information without projection, the neighbor information is less compressed. 2) voxels are labeled and binarized according to the buffer-weighted density, which improves the robustness to boundary effect, noise and density variation. Based on the VBBD, a global optimal surface matching method based on a Kuhn_Munkres (KM) algorithm is adopted. The strategy has several advantages: 1) by calculating the descriptors of large-scaled surfaces which are down sampled, the number of the point cloud is reduced, and meanwhile, large-scaled information is obtained; and 2) KM algorithm is adopted to find matching pairs to achieve final maximum weight sum of all matching pairs, which can efficiently avoid local optimum. The experimental results show that, compared with other state-of-the-art descriptors, the VBBD has better descriptiveness and robustness, and the surface matching strategy by the VBBD can achieve both high recall and precision.

Extended Local Binary Patterns for Efficient and Robust Spontaneous Facial Micro-Expression Recognition

Facial Micro-Expressions (MEs) are spontaneous, involuntary facial movements when a person experiences an emotion but deliberately or unconsciously attempts to conceal his or her genuine emotions. Recently, ME recognition has attracted increasing attention due to its potential applications such as clinical diagnosis, business negotiation, interrogations, and security. However, it is expensive to build large scale ME datasets, mainly due to the difficulty of inducing spontaneous MEs. This limits the application of deep learning techniques which require lots of training data. In this paper, we propose a simple, efficient yet robust descriptor called Extended Local Binary Patterns on Three Orthogonal Planes (ELBPTOP) for ME recognition. ELBPTOP consists of three complementary binary descriptors: LBPTOP and two novel ones Radial Difference LBPTOP (RDLBPTOP) and Angular Difference LBPTOP (ADLBPTOP), which explore the local second order information along the radial and angular directions contained in ME video sequences. ELBPTOP is a novel ME descriptor inspired by unique and subtle facial movements. It is computationally efficient and only marginally increases the cost of computing LBPTOP, yet is extremely effective for ME recognition. In addition, by firstly introducing Whitened Principal Component Analysis (WPCA) to ME recognition, we can further obtain more compact and discriminative feature representations, then achieve significantly computational savings. Extensive experimental evaluation on three popular spontaneous ME datasets SMIC, CASME II and SAMM show that our proposed ELBPTOP approach significantly outperforms the previous state-of-the-art on all three single evaluated datasets and achieves promising results on cross-database recognition.Our code will be made available.

DETERMINATION OF RELEVANCE OF VISUAL OBJECT IMAGES BY APPLICATION OF STATISTICAL ANALYSIS OF REGARDING FRAGMENT REPRESENTATION OF THEIR DESCRIPTIONS

The problem of visual object recognition in the space features of descriptors of special image points is solved with the application of statistical distributions for a system of fragments of image description. The efficiency of application of the distribution variants for single- and multi-bit sizes of the fragments of binary descriptors in the calculation of the level of relevance of descriptions in the process of object recognition is analyzed. Using software simulation, an experimental evaluation of the computation time of the relevance value in the constructed feature space is carried out compared to the traditional voting approach of the set of descriptors. The efficiency of the proposed approach to data representation in terms of a significant increase in the performance rate with ensuring a sufficient level of recognition quality is confirmed.

Impact of reduction in descriptor size on object detection and classification

Extraction of distinctive and robust features in image/video analysis and processing has attracted the attention of researchers in the recent years. Elimination of irrelevant and less important features reduces the computational complexity to a great extent at the cost of a very marginal reduction in accuracy. This paper presents a framework for dimensionality reduction of the binary features to obtain a low dimension feature vector for object detection. The process of identification and selection of the most relevant feature is performed in three steps: extraction of features using binary descriptors; Selection of best feature subset using Sequential Forward Selection (SFS) and Principal Component Analysis; classification using SVM classifier. The experimental results show that BRISK and LATCH descriptors perform better even when the dimensionality is reduced from 256, 128, and 64 to 32 bits with an acceptable classification accuracy and significant reduction in run time. However, there is slight decrease in the classification accuracy of LBP, FREAK, BRIEF, and ORB. A classification rate of 84.93% is obtained with LATCH descriptor for a descriptor size of 32 bits.

Combining MRF-based deformable registration and deep binary 3D-CNN descriptors for large lung motion estimation in COPD patients.

Deep convolutional neural networks in their various forms are currently achieving or outperforming state-of-the-art results on several medical imaging tasks. We aim to make these developments available to the so far unsolved task of accurate correspondence finding-especially with regard to image registration. We propose a two-step hybrid approach to make deep learned features accessible to a discrete optimization-based registration method. In a first step, in order to extract expressive binary local descriptors, we train a deep network architecture on a patch-based landmark retrieval problem as auxiliary task. As second step at runtime within a MRF-regularised dense displacement sampling, their binary nature enables highly efficient similarity computations, thus making them an ideal candidate to replace the so far used handcrafted local feature descriptors during the registration process. We evaluate our approach on finding correspondences between highly non-rigidly deformed lung CT scans from different breathing states. Although the CNN-based descriptors excell at an auxiliary learning task for finding keypoint correspondences, self-similarity-based descriptors yield more accurate registration results. However, a combination of both approaches turns out to generate the most robust features for registration. We present a three-dimensional framework for large lung motion estimation based on the combination of CNN-based and handcrafted descriptors efficiently employed in a discrete registration method. Achieving best results by combining learned and handcrafted features encourages further research in this direction.

Copy-move forgery detection using binary discriminant features

Digital images have numerous applications in various fields such as cyber forensics, courts, newspapers, magazines, medical field and many more. Therefore, the trustfulness of the contents of an image is very essential. Image manipulation has become very much simple and common due to the development of high resolution digital cameras, powerful computers and numerous image editing software. Hence, falsification of visual content of images is not at all limited to specialists and detection of forged images have become a highly challenging task. Of the different kinds of image forgery, the proposed method deals with one particular kind known as copy-move forgery. The method is an integration of the conventional block-based and keypoint based methods. The method first identifies the forged keypoint locations and then localizes the forged region using a region extraction technique. To improve the accuracy of detection, a Binary Discriminative Feature descriptor (BDF) is used for feature detection and matching. The suspected regions are identified by replacing the matching feature points with corresponding superpixel blocks. The neighbouring blocks are then merged with the suspected regions based on color similarity, using Color Histogram Matching and a final morphological close operation extracts the detected forged regions. Experimental results show that the proposed method has better detection accuracy, in terms of precision, recall and F1 Score, when compared with the state-of-the-art methods of copy-move forgery detection, under conditions of plain copy-move forgery as well as post-processing conditions like brightness changes, contrast adjustments, color reduction and blurring.

Human ear recognition based on local multi-scale LBP features with city-block distance

The use of the ear as a biometric modality has emerged in recent years. It makes it possible to differentiate people thanks to its stability over time and to the richness of its characteristics such as texture, color and size. This paper proposes a novel approach to ear recognition based on a variant of the Local Binary Pattern descriptor called Multi-scale Local Binary Pattern (MLBP). MLBP is calculated locally, by dividing the image into several equal blocks, to extract the ear features which will be used in the matching process to make a decision by detecting the similarities between the feature vectors using City-Block distance (CTB). The proposed method is evaluated on three reference ear databases: IIT Delhi I, IIT Delhi II and USTB-1. The analysis of the results obtained have clearly shown the robustness and the stability of the proposed ear recognition method which is highly competitive, achieving an attractive recognition performances in terms of identification rate at rank-1 up to: 98.40% for IIT Delhi I, 98.64% for IIT Delhi II, and 98.33% for USTB-1.

Two improved extension of local binary pattern descriptors using wavelet transform for texture classification

Texture image analysis plays a pivotal role in pattern recognition and image retrieval. In this study, two improved local binary pattern descriptors are proposed using wavelet transform for texture analysis. The two proposed methods, namely wavelet domain local statistical binary pattern (WLSBP) and directional WLSBP (dWLSBP α ) both consist of three stages. In the first stage, discrete wavelet decomposition is applied to decompose the image. In the second stage, the proposed statistical parameters are computed from the decomposed image, which results in binary value 0 or 1. Then, the binary values are transformed into WLSBP/dWLSBP α label. In the third stage, the histogram is built using the WLSBP/dWLSBP α labels. The proposed WLSBP and dWLSBP α differ in terms of considering the neighbours. The proposed WLSBP considers the neighbours circularly, whereas dWLSBP α considers the neighbours in the same orientation through the central wavelet coefficient. The proposed approaches have been applied for copy-move forgery detection. Experiments show that the performance of the proposed methods has improved retrieval rate compared with existing methods on both Brodatz and Outex databases.

A binary method of multisensor image registration based on angle traversal

Registration is a key step in image preprocessing. Absorbing the binary idea of BRIEF, we took all the feature points as a whole and generated a new binary descriptor that was irrelevant to the gray of the neighborhood of a feature point in the registration of visible light with infrared images. For each selected feature point, we constructed a circular area centered on it and divided the circumference equally as the sampling interval of the feature points. We then traversed all the intervals and built a binary code string as a descriptor for the feature point according to the relative position of the feature point. During the process of angle traversal, the Hamming distance was applied to identify the most similar descriptor pair. We then adopted an improved clustering algorithm to eliminate mismatched pairs. Experiments have demonstrated that this kind of descriptor has high accuracy and robustness for multisensor remote-sensing images. In particular, when dealing with the combination of visible and far-infrared images with local grayscale inversion, registration can also be accomplished well.

Facial expression recognition using best tree RD-LGP encoded features and HMM

Hand and face gestures enable deaf people to communicate in their daily lives rather than speaking. This paper describes a hidden Markov model (HMM)-based framework for face sign recognition and detection. The observation vectors used to characterize the states of the HMM are obtained using the best tree local gradient pattern (LGP) encoded features. Each face gesture is modeled as a five-state HMM. The problem of facial expression classification is posed as a composite seven-classes multi-hypothesis Bayesian test. The likelihood ratio test showed that the overall recognition rate for the proposed model is higher than the HMM-local binary pattern descriptor by 6.4%. The overall recognition rate is enhanced by 8.6% using the discrete wavelet packet best tree decomposition filter as a pre-processing noise removal tool. In addition, the overall recognition rate ranges from 84.3%, for the seven classes Bayesian test, to 100%, for lower number of classes depending upon the type of the face gesture. The proposed face expression algorithm reduces significantly the computational complexity of previous HMM-based face expression recognition systems, and still preserve the recognition rate.

Local Binary Pattern Family Descriptors for Texture Classification

Local Binary Pattern Family Descriptors for Texture Classification

HBST: A Hamming Distance Embedding Binary Search Tree for Feature-Based Visual Place Recognition

Reliable and efficient Visual Place Recognition is a major building block of modern SLAM systems. Leveraging on our prior work, in this paper we present a Hamming Distance embedding Binary Search Tree (HBST) approach for binary Descriptor Matching and Image Retrieval. HBST allows for descriptor Search and Insertion in logarithmic time by exploiting particular properties of binary Feature descriptors. We support the idea behind our search structure with a thorough analysis on the exploited descriptor properties and their effects on completeness and complexity of search and insertion. To validate our claims we conducted comparative experiments for HBST and several state-of-the-art methods on a broad range of publicly available datasets. HBST is available as a compact open-source C++ header-only library.

IBoW-LCD: An Appearance-Based Loop-Closure Detection Approach Using Incremental Bags of Binary Words

In this paper, we introduce iBoW-LCD, a novel appearance-based loop closure detection method. The presented approach makes use of an incremental Bag-of-Words (BoW) scheme based on binary descriptors to retrieve previously seen similar images, avoiding any vocabulary training stage usually required by classic BoW models. In addition, to detect loop closures, iBoW-LCD builds on the concept of dynamic islands, a simple but effective mechanism to group similar images close in time, which reduces the computational times typically associated to Bayesian frameworks. Our approach is validated using several indoor and outdoor public datasets, taken under different environmental conditions, achieving a high accuracy and outperforming other state-of-the-art solutions.