scholarly journals Study on the Influence of Image Noise on Monocular Feature-Based Visual SLAM Based on FFDNet

Sensors ◽  
2020 ◽  
Vol 20 (17) ◽  
pp. 4922
Author(s):  
Like Cao ◽  
Jie Ling ◽  
Xiaohui Xiao
Keyword(s):  
Visual Slam ◽  
Scale Invariant ◽  
Original Sequence ◽  
Speeded Up Robust Features ◽  
Localization And Mapping ◽  
Feature Based ◽  
Scale Invariant Feature ◽  
Influence Of Noise ◽  
Denoised Image ◽  
Better Than

Noise appears in images captured by real cameras. This paper studies the influence of noise on monocular feature-based visual Simultaneous Localization and Mapping (SLAM). First, an open-source synthetic dataset with different noise levels is introduced in this paper. Then the images in the dataset are denoised using the Fast and Flexible Denoising convolutional neural Network (FFDNet); the matching performances of Scale Invariant Feature Transform (SIFT), Speeded Up Robust Features (SURF) and Oriented FAST and Rotated BRIEF (ORB) which are commonly used in feature-based SLAM are analyzed in comparison and the results show that ORB has a higher correct matching rate than that of SIFT and SURF, the denoised images have a higher correct matching rate than noisy images. Next, the Absolute Trajectory Error (ATE) of noisy and denoised sequences are evaluated on ORB-SLAM2 and the results show that the denoised sequences perform better than the noisy sequences at any noise level. Finally, the completely clean sequence in the dataset and the sequences in the KITTI dataset are denoised and compared with the original sequence through comprehensive experiments. For the clean sequence, the Root-Mean-Square Error (RMSE) of ATE after denoising has decreased by 16.75%; for KITTI sequences, 7 out of 10 sequences have lower RMSE than the original sequences. The results show that the denoised image can achieve higher accuracy in the monocular feature-based visual SLAM under certain conditions.

scholarly journals Dog Identification Method Based on Muzzle Pattern Image

2020 ◽  
Vol 10 (24) ◽  
pp. 8994
Author(s):  
Dong-Hwa Jang ◽  
Kyeong-Seok Kwon ◽  
Jung-Kon Kim ◽  
Ka-Young Yang ◽  
Jong-Bok Kim
Keyword(s):  
Feature Extraction ◽  
Radio Frequency Identification ◽  
Hamming Distance ◽  
Histogram Equalization ◽  
Scale Invariant ◽  
Speeded Up Robust Features ◽  
Identification Method ◽  
Approximate Nearest Neighbors ◽  
Frequency Identification ◽  
Scale Invariant Feature

Currently, invasive and external radio frequency identification (RFID) devices and pet tags are widely used for dog identification. However, social problems such as abandoning and losing dogs are constantly increasing. A more effective alternative to the existing identification method is required and the biometrics can be the alternative. This paper proposes an effective dog muzzle recognition method to identify individual dogs. The proposed method consists of preprocessing, feature extraction, matching, and postprocessing. For preprocessing, proposed resize and histogram equalization are used. For feature extraction algorithm, Scale Invariant Feature Transform (SIFT), Speeded Up Robust Features (SURF), Binary Robust Invariant Scaling Keypoints (BRISK) and Oriented FAST, and Rotated BRIEF (ORB) are applied and compared. For matching, Fast Library for Approximate Nearest Neighbors (FLANN) is used for SIFT and SURF, and hamming distance are used for BRISK and ORB. For postprocessing, two techniques to reduce incorrect matches are proposed. The proposed method was evaluated with 55 dog muzzle pattern images acquired from 11 dogs and 990 images augmented by the image deformation (i.e., angle, illumination, noise, affine transform). The best Equal Error Rate (EER) of the proposed method was 0.35%, and ORB was the most appropriate for the dog muzzle pattern recognition.

scholarly journals Analysis of Application of Cluster Descriptions in Space of Characteristic Image Features

Data ◽  
2018 ◽  
Vol 3 (4) ◽  
pp. 52 ◽  
Author(s):  
Oleksii Gorokhovatskyi ◽  
Volodymyr Gorokhovatskyi ◽  
Olena Peredrii
Keyword(s):  
Recognition Performance ◽  
Input Image ◽  
Image Features ◽  
Additional Time ◽  
Scale Invariant ◽  
Speeded Up Robust Features ◽  
Image Descriptors ◽  
Structural Image ◽  
Characteristic Features ◽  
Scale Invariant Feature

In this paper, we propose an investigation of the properties of structural image recognition methods in the cluster space of characteristic features. Recognition, which is based on key point descriptors like SIFT (Scale-invariant Feature Transform), SURF (Speeded Up Robust Features), ORB (Oriented FAST and Rotated BRIEF), etc., often relating to the search for corresponding descriptor values between an input image and all etalon images, which require many operations and time. Recognition of the previously quantized (clustered) sets of descriptor features is described. Clustering is performed across the complete set of etalon image descriptors and followed by screening, which allows for representation of each etalon image in vector form as a distribution of clusters. Due to such representations, the number of computation and comparison procedures, which are the core of the recognition process, might be reduced tens of times. Respectively, the preprocessing stage takes additional time for clustering. The implementation of the proposed approach was tested on the Leeds Butterfly dataset. The dependence of cluster amount on recognition performance and processing time was investigated. It was proven that recognition may be performed up to nine times faster with only a moderate decrease in quality recognition compared to searching for correspondences between all existing descriptors in etalon images and input one without quantization.

Image-Searching for Office Equipment Using Bag-of-Keypoints and AdaBoost

2011 ◽  
Vol 23 (6) ◽  
pp. 1080-1090 ◽  
Author(s):  
Seiji Aoyagi ◽  
◽  
Atsushi Kohama ◽  
Yuki Inaura ◽  
Masato Suzuki ◽  
...  
Keyword(s):  
Input Image ◽  
Image Features ◽  
Scale Invariant ◽  
Edge Information ◽  
Localization And Mapping ◽  
Template Image ◽  
Specific Object ◽  
Monocular Image ◽  
Scale Invariant Feature ◽  
Office Equipment

For an indoor mobile robot’s Simultaneous Localization And Mapping (SLAM), a method of processing only one monocular image (640×480 pixel) of the environment is proposed. This method imitates a human’s ability to grasp at a glance the overall situation of a room, i.e., its layout and any objects or obstacles in it. Specific object recognition of a desk through the use of several camera angles is dealt with as one example. The proposed method has the following steps. 1) The bag-of-keypoints method is applied to the image to detect the existence of the object in the input image. 2) If the existence of the object is verified, the angle of the object is further detected using the bag-ofkeypoints method. 3) The candidates for the projection from template image to input image are obtained using Scale Invariant Feature Transform (SIFT) or edge information. Whether or not the projected area correctly corresponds to the object is checked using the AdaBoost classifier, based on various image features such as Haar-like features. Through these steps, the desk is eventually extractedwith angle information if it exists in the image.

A SIFT Feature-Based Template Matching Method for Detecting and Counting Objects in Life Space

2010 ◽  
Vol 36 ◽  
pp. 413-421 ◽  
Author(s):  
Hideaki Kawano ◽  
Hideaki Orii ◽  
Katsuaki Shiraishi ◽  
Hiroshi Maeda
Keyword(s):  
Template Matching ◽  
Autonomous Robots ◽  
Target Object ◽  
Main Process ◽  
Spatial Transformation ◽  
Scale Invariant ◽  
Everyday Objects ◽  
Counting Task ◽  
Feature Based ◽  
Scale Invariant Feature

Autonomous robots are at advanced stage in various fields, and they are expected to autonomously work at the scenes of nursing care or medical care in the near future. In this paper, we focus on object counting task by images. Since the number of objects is not a mere physical quantity, it is difficult for conventional phisical sensors to measure such quantity and an intelligent sensing with higher-order recognition is required to accomplish such counting task. It is often that we count the number of objects in various situations. In the case of several objects, we can recognize the number at a glance. On the other hand, in the case of a dozen of objects, the task to count the number might become troublesome. Thus, simple and easy way to enumerate the objects automatically has been expected. In this study, we propose a method to recognize the number of objects by image. In general, the target object to count varies according to user's request. In order to accept the user's various requests, the region belonging to the desired object in the image is selected as a template. Main process of the proposed method is to search and count regions which resembles the template. To achieve robustness against spatial transformation, such as translation, rotation, and scaling, scale-invariant feature transform (SIFT) is employed as a feature. To show the effectiveness, the proposed method is applied to few images containing everyday objects, e.g., binders, cans etc.

Comparison of Contour Based and Feature Based Tracking Methods for Control of Microbiorobots

2009 ◽  
Author(s):  
Dal Hyung Kim ◽  
Edward Steager ◽  
Min Jun Kim
Keyword(s):  
Small Error ◽  
Small Scale ◽  
Scale Invariant ◽  
Sift Algorithm ◽  
Tracking Method ◽  
Miniature Robots ◽  
A Cell ◽  
Feature Based ◽  
Scale Invariant Feature ◽  
Exact Positions

Miniature robots should be precisely controlled because of a small workspace and size of their shapes. Small error of control could lead to failure of tasks such as an assembly. Tracking is one of the most important techniques because control of a small scale robot is hard to accomplish without object’s motion information. In this paper, we compare the feature based and the region based tracking methods with microbiorobot. Invariant features can be extracted using Scale Invariant Feature Transfrom (SIFT) algorithm because microbiorobot is a rigid body unlike a cell. We clearly showed that the feature based tracking method track exact positions of the objects than region based tracking method when objects are close contacted or overlapped. Also, the feature based tracking method allows tracking of objects even though partial object disappears or illumination is changed.

Feature-Based Real-Time Visual SLAM Using Kinect

2014 ◽  
Vol 989-994 ◽  
pp. 2651-2654
Author(s):  
Yan Song ◽  
Bo He
Keyword(s):  
Real Time ◽  
3D Models ◽  
Image Features ◽  
Microsoft Kinect ◽  
Depth Information ◽  
Visual Slam ◽  
Depth Data ◽  
Localization And Mapping ◽  
Speed Up ◽  
Feature Based

In this paper, a novel feature-based real-time visual Simultaneous localization and mapping (SLAM) system is proposed. This system generates colored 3-D reconstruction models and 3-D estimated trajectory using a Kinect style camera. Microsoft Kinect, a low priced 3-D camera, is the only sensor we use in our experiment. Kinect style sensors give RGB-D (red-green-blue depth) data which contains 2D image and per-pixel depth information. ORB (Oriented FAST and Rotated BRIEF) is the algorithm used to extract image features for speed up the whole system. Our system can be used to generate 3-D detailed reconstruction models. Furthermore, an estimated 3D trajectory of the sensor is given in this paper. The results of the experiments demonstrate that our system performs robustly and effectively in both getting detailed 3D models and mapping camera trajectory.

scholarly journals Improved Point-Line Feature Based Visual SLAM Method for Indoor Scenes

Sensors ◽  
2018 ◽  
Vol 18 (10) ◽  
pp. 3559 ◽  
Author(s):  
Runzhi Wang ◽  
Kaichang Di ◽  
Wenhui Wan ◽  
Yongkang Wang
Keyword(s):  
Line Segment ◽  
Hessian Matrix ◽  
Visual Slam ◽  
Localization Accuracy ◽  
Stereo Camera ◽  
Line Feature ◽  
Localization And Mapping ◽  
Indoor Scenes ◽  
Feature Based ◽  
Point Line

In the study of indoor simultaneous localization and mapping (SLAM) problems using a stereo camera, two types of primary features—point and line segments—have been widely used to calculate the pose of the camera. However, many feature-based SLAM systems are not robust when the camera moves sharply or turns too quickly. In this paper, an improved indoor visual SLAM method to better utilize the advantages of point and line segment features and achieve robust results in difficult environments is proposed. First, point and line segment features are automatically extracted and matched to build two kinds of projection models. Subsequently, for the optimization problem of line segment features, we add minimization of angle observation in addition to the traditional re-projection error of endpoints. Finally, our model of motion estimation, which is adaptive to the motion state of the camera, is applied to build a new combinational Hessian matrix and gradient vector for iterated pose estimation. Furthermore, our proposal has been tested on EuRoC MAV datasets and sequence images captured with our stereo camera. The experimental results demonstrate the effectiveness of our improved point-line feature based visual SLAM method in improving localization accuracy when the camera moves with rapid rotation or violent fluctuation.

Sign in / Sign up

Export Citation Format

Share Document