Feature points detection and tracking based on SIFT combining with KLT method

To explore the impact of different image registration algorithms on the diagnosis of visual path damage in patients with primary open angle glaucoma (POAG), 60 cases of suspected POAG patients were selected as the research objects. Shape-preserving scale invariant feature transform (SP-SIFT) algorithm, scale invariant feature transform (SIFT) algorithm, and Kanade-Lucas-Tomasi (KLT) algorithm were compared and applied to MRI images of 60 POAG patients. It was found that the SP-SIFT algorithm converged after 33 iterations, which had a higher registration speed than the SIFT algorithm and the KLT algorithm. The mean errors of the SP-SIFT algorithm in the rotation angle, X-direction translation, and Y-direction translation were 2.11%, 4.56%, and 4.31%, respectively. Those of the SIFT algorithm were 5.55%, 9.98%, and 7.01%, respectively. Those of the KLT algorithm were 7.45%, 11.31%, and 8.56%, respectively, and the difference among algorithms was significant ( P < 0.05 ). The diagnostic sensitivity and accuracy of the SP-SIFT algorithm for POAG were 96.15% and 94.34%, respectively. Those of the SIFT algorithm were 94.68% and 90.74%, respectively. Those of the KLT algorithm were 94.21% and 90.57%, respectively, and the three algorithms had significant differences ( P < 0.05 ). The results of MRI images based on the SP-SIFT algorithm showed that the average thickness of the cortex of the patient’s left talar sulcus, right talar sulcus, left middle temporal gyrus, and left fusiform gyrus were 2.49 ± 0.15 mm, 2.62 ± 0.13 mm, 3.00 ± 0.10 mm, and 2.99 ± 0.17 mm, respectively. Those of the SIFT algorithm were 2.51 ± 0.17 mm, 2.69 ± 0.12 mm, 3.11 ± 0.13 mm, and 3.09 ± 0.14 mm, respectively. Those of the KLT algorithm were 2.35 ± 0.12 mm, 2.52 ± 0.16 mm, 2.77 ± 0.11 mm, and 2.87 ± 0.17 mm, respectively, and the three algorithms had significant differences ( P < 0.05 ). In summary, the SP-SIFT algorithm was ideal for POAG visual pathway diagnosis and was of great adoption potential in clinical diagnosis.

Feature Point (FP) detection is a fundamental step in computer vision tasks. Although FP detectors are mostly designed to support Low Dynamic Range (LDR) images as input, interest in High Dynamic Range (HDR) images has increased recently due to their higher precision to register overexposed and underexposed areas in an image. As the detection of FPs is strongly dependent on the illumination of the scene, HDR images have the potential to be more robust than LDR images during FP detection. Known FP detectors, however, do not use the full potential of HDR images. In addition, few works have evaluated the performance of HDR images in this context. In this paper, we propose a modification of FP detectors aiming to improve FP detection on HDR images. To this end, we design a local mask based on the Coefficient of Variation (CV) of sets of pixels, creating thus a new step in FP detection. We compare our approach with popular FP detection methods using a standard evaluation metric, Repeatability Rate (RR) of FPs, and also Uniformity as a proposed new criterion. A dataset of images from scenes affected by camera transformations and substantial illumination changes was used as input. Experimental results show that our proposed algorithms give better Uniformity and RR in most HDR images from the dataset when compared to standard FP detectors. Moreover, they indicate that HDR images present a great potential to be explored in applications that rely on FP detection.

Road visual navigation relies on accurate road models. This study was aimed at proposing an improved scale-invariant feature transform (SIFT) algorithm for recovering depth information from farmland road images, which would provide a reliable path for visual navigation. The mean image of pixel value in five channels (R, G, B, S and V) were treated as the inspected image and the feature points of the inspected image were extracted by the Canny algorithm, for achieving precise location of the feature points and ensuring the uniformity and density of the feature points. The mean value of the pixels in 5×5 neighborhood around the feature point at an interval of 45o in eight directions was then treated as the feature vector, and the differences of the feature vectors were calculated for preliminary matching of the left and right image feature points. In order to achieve the depth information of farmland road images, the energy method of feature points was used for eliminating the mismatched points. Experiments with a binocular stereo vision system were conducted and the results showed that the matching accuracy and time consuming for depth recovery when using the improved SIFT algorithm were 96.48% and 5.6 s, respectively, with the accuracy for depth recovery of –7.17%-2.97% in a certain sight distance. The mean uniformity, time consuming and matching accuracy for all the 60 images under various climates and road conditions were 50%-70%, 5.0-6.5 s, and higher than 88%, respectively, indicating that performance for achieving the feature points (e.g., uniformity, matching accuracy, and algorithm real-time) of the improved SIFT algorithm were superior to that of conventional SIFT algorithm. This study provides an important reference for navigation technology of agricultural equipment based on machine vision. Keywords: scale-invariant feature transform (sift), feature matching, canny operator, energy method of feature point, farmland road, depth recovery, visual navigation DOI: 10.25165/j.ijabe.20191204.4821 Citation: Yao L J, Hu D, Yang Z D, Li H B, Qian M B. Depth recovery for unstructured farmland road image using an improved SIFT algorithm. Int J Agric & Biol Eng, 2019; 12(4): 141–147.

A novel digital image stabilization approach using Harris and Scale Invariant Feature Transform (SIFT) was presented in this article. Using SIFT in digital image stabilization, too many feature points and matches were extracted, but some of them were not so stable. Using these feature points and matches can not only increase the computational effort, but also enhance the wrong matching probability. We proposed to use SIFT to detect feature points and incorporate the Harris criterion to select the most stable feature points in the video sequence where image motion was happened due to vehicle or platform vibration. With these feature points, we use general feature descriptor and matching algorithm to achieve the image stabilization. Experimental results show that the proposed algorithm can not only bring down the probability of wrong matching and get more accurate matches, but also reduce the computation burden than SIFT effectively.

A novel scale-invariant feature transform (SIFT) algorithm is proposed for soccer target recognition application in a robot soccer game. First, the method of generating scale space is given, extreme points are detected. This gives the precise positioning of the extraction step and the SIFT feature points. Based on the gradient and direction of the feature point neighboring pixels, a description of the key points of the vector is generated. Finally, the matching method based on feature vectors is extracted from SIFT feature points and implemented on the image of the football in a soccer game. By employing the proposed SIFT algorithm for football and stadium key feature points extraction and matching, significant increase can be achieved in the robot soccer ability to identify and locate the football.

This paper presents an unsupervised learning-based method for selection of feature points and object category classification without previous setting of the number of categories. Our method consists of the following procedures: 1) detection of feature points and description of features using a Scale-Invariant Feature Transform (SIFT), 2) selection of target feature points using One Class-Support Vector Machines (OC-SVMs), 3) generation of visual words of all SIFT descriptors and histograms in each image of selected feature points using Self-Organizing Maps (SOMs), 4) formation of labels using Adaptive Resonance Theory-2 (ART-2), and 5) creation and classification of categories on a category map of Counter Propagation Networks (CPNs) for visualizing spatial relations between categories. Classification results of static images using a Caltech-256 object category dataset and dynamic images using time-series images obtained using a robot according to movements respectively demonstrate that our method can visualize spatial relations of categories while maintaining time-series characteristics. Moreover, we emphasize the effectiveness of our method for category classification of appearance changes of objects.

High speed robust image registration and localization using optimized algorithm and its performances evaluation

얼굴추적은 3차원 공간상에서 머리(head)와 안면(face)의 움직임을 추정하는 기술로, 얼굴 표정 감정인식과 같은 상위 분석단계의 중요한 기반기술이다. 본 논문에서는 AAM 기반의 얼굴추적 알고리즘을 제안한다. AAM은 변형되는 대상을 분할하고 추적하는데 광범위하게 적용되고 있다. 그러나 여전히 여러 가지 해결해야할 제약사항들이 있다. 특히 자체중첩(self-occlusion)과 부분적인 중첩, 그리고 일시적으로 완전히 가려지는 완전중첩 상황에서 보통 국부해에 수렴(local convergence)하거나 발산하기 쉽다. 본 논문에서는 이러한 중첩상황에 대한 AAM의 강인성을 향상시키기 위해서 SIFT 특징을 이용하고 있다. SIFT는 일부 영상의 특징점으로 안정적인 추적이 가능하기 때문에 자체와 부분중첩에 효과적이며, 완전중첩의 상황에도 SIFT의 전역적인 매칭성능으로 별도의 재초기화 없이 연속적인 추적이 가능하다. 또한 추적과정에서 큰 자세변화에 따른 움직임을 효과적으로 추정하기 위해서 다시점(multi-view) 얼굴영상의 SIFT 특징을 온라인으로 등록하여 활용하고 있다. 제안한 알고리즘의 이러한 강인성은 위 세 가지 중첩상황에 대해서 기존 알고리즘들과의 비교실험을 통해서 보여준다. Face tracking is to estimate the motion of a non-rigid face together with a rigid head in 3D, and plays important roles in higher levels such as face/facial expression/emotion recognition. In this paper, we propose an AAM-based face tracking algorithm. AAM has been widely used to segment and track deformable objects, but there are still many difficulties. Particularly, it often tends to diverge or converge into local minima when a target object is self-occluded, partially or completely occluded. To address this problem, we utilize the scale invariant feature transform (SIFT). SIFT is an effective method for self and partial occlusion because it is able to find correspondence between feature points under partial loss. And it enables an AAM to continue to track without re-initialization in complete occlusions thanks to the good performance of global matching. We also register and use the SIFT features extracted from multi-view face images during tracking to effectively track a face across large pose changes. Our proposed algorithm is validated by comparing other algorithms under the above 3 kinds of occlusions.

Feature point detection is an important pre-processing step for quantitative evaluation of facial paralysis. Since the conventional methods such as active shape model (ASM) or active appearance model (AAM) are trained by using normal face and they are not possible to detect the feature points accurately for the face with paralysis. In this paper, we propose an automatic and accurate feature point detection method for quantitative evaluation of facial paralysis using deep convolutional neural networks (DCNN). The proposed method consists of two steps. We first use AAM for initial feature point detection. In the second step, a patch with the detected point at the center is used as an input of DCNN for refinement. Experiments demonstrated that the proposed method can significantly improve the detection accuracy of the conventional AAM.

A novel image registration method is proposed based on the scale-invariant feature transform (SIFT). Our approach combines the image frequency spectrum with a regular grid. The method exploits features of SIFT, the greater number of feature points in high-frequency areas of an image and regular grid traits with a fine spatial distribution to extract evenly distributed feature points. Global correspondence error checking is used to remove false matching pairs. The registration of eight groups of remote sensing images shows that our method has well-distributed feature points in the spatial and scale domains, obtains an average of 15.78% matching pairs more than uniform robust (UR)-SIFT, and provides better registration accuracy than SIFT and UR-SIFT.

There remain many difficult problems in computer vision research such as object recognition, three dimensional reconstruction, object tracking, etc. And the basis of solving these problems relies on image matching. The scale invariant feature transform (SIFT) algorithm has been widely used for image matching application. The SIFT algorithm can successfully extract the most descriptive feature points in given input images taken under different viewpoints. However, the performance of the original SIFT algorithm degrades under the influence of noise. We propose to modify the SIFT algorithm to produce better invariant feature points for image matching under noise. We also propose to employ the Earth mover's distance (EMD) as the measurement of similarity between two descriptors. We present extensive experiment results to demonstrate the performance of the proposed methods in image matching under noise.

In this paper, we propose a semi-automatic object tracking method based on a Scale Invariant Feature Transform (SIFT) and Kanade-Lucas-Tomasi (KLT) tracker. In our approach, the region of interest is specified by the user and then the interest points are detected. The tracker is then used to track the specified object in the consecutive frames. To overcome rapid changes of appearance, occlusion or disappearance from the camera view, we employ a forward-backward error compensation. Experimental results on VIVID dataset indicates that the proposed method has superior overall performance compared to more common methods in the field.

Feature-based retinal fundus image registration (RIR) technique aligns fundus images according to geometrical transformations estimated between feature point correspondences. To ensure accurate registration, the feature points extracted must be from the retinal vessels and throughout the image. However, noises in the fundus image may resemble retinal vessels in local patches. Therefore, this paper introduces a feature extraction method based on a local feature of retinal vessels (CURVE) that incorporates retinal vessels and noises characteristics to accurately extract feature points on retinal vessels and throughout the fundus image. The CURVE performance is tested on CHASE, DRIVE, HRF and STARE datasets and compared with six feature extraction methods used in the existing feature-based RIR techniques. From the experiment, the feature extraction accuracy of CURVE (86.021%) significantly outperformed the existing feature extraction methods (p ≤ 0.001*). Then, CURVE is paired with a scale-invariant feature transform (SIFT) descriptor to test its registration capability on the fundus image registration (FIRE) dataset. Overall, CURVE-SIFT successfully registered 44.030% of the image pairs while the existing feature-based RIR techniques (GDB-ICP, Harris-PIIFD, Ghassabi’s-SIFT, H-M 16, H-M 17 and D-Saddle-HOG) only registered less than 27.612% of the image pairs. The one-way ANOVA analysis showed that CURVE-SIFT significantly outperformed GDB-ICP (p = 0.007*), Harris-PIIFD, Ghassabi’s-SIFT, H-M 16, H-M 17 and D-Saddle-HOG (p ≤ 0.001*).

Simultaneous localization and mapping (SLAM) is a key research content of robot autonomous navigation, the visual monocular SLAM based on Extend Kalman Filter(EKF) is one important method to handle this problem. But due to high computational complexity, it has strict limits on the number and stability of the feature points, traditional method selects few corners like or straight lines as feature points, and these methods limit the application scope of EKF-SLAM. This paper proposes a key points selection method based on SIFT(Scale-invariant feature transform) feature point, on the assumption of relative uniform of the feature points' distribution, through controlling the total number of feature points effectively, the applied restriction of the visual monocular EKF-SLAM is reduced. Experiments show that this feature point selection method has a high stability for different scenes, and improves the convergence velocity.

This paper presents a method to extract 3D outlines of objects in an image obtained from a monocular vision. After detecting the general outlines of the object by MOPS(Multi-Scale Oriented Patches) -algorithm and we obtain their spatial coordinates. Simultaneously, it obtains the space-coordinates with feature points to be immanent within the outlines of objects through SIFT(Scale Invariant Feature Transform)-algorithm. It grasps a form of objects to join the space-coordinates of outlines and SIFT feature points. The method which is proposed in this paper, it forms general outlines of objects, so that it enables a rapid calculation, and also it has the advantage capable of collecting a detailed data because it supplies the internal-data of outlines through SIFT feature points.