- 标题
- 摘要
- 关键词
- 实验方案
- 产品
-
New approach to enhancing the performance of cloud-based vision system of mobile robots
摘要： Mobile robots require real-time performance, high computation power, and a shared computing environment. Although cloud computing offers computation power, it may adversely affect real-time performance owing to network lag. The main objective of this study is to allow a mobile robot vision system to reliably achieve real-time constraints using cloud computing. A human cloud mobile robot architecture is proposed as well as a data flow mechanism organized on both the mobile robot and the cloud server sides. Two algorithms are proposed: (i) A real-time image clustering algorithm, applied on the mobile robot side, and (ii) A modified growing neural gas algorithm, applied on the cloud server side. The experimental results demonstrate that there is a 25% to 45% enhancement in the total response time, depending on the communication bandwidth and image resolution. Moreover, better performance in terms of data size, path planning time, and accuracy is demonstrated over other state-of-the-art techniques.
关键词： Computation offloading,Computer vision,3D point cloud,Mobile robot,Stereo vision,Real-time networking,Cloud computing,Cloud robotics
更新于2025-09-23 15:23:52
-
Ground based hyperspectral imaging for extensive mango yield estimation
摘要： Fruit yield estimation in orchard blocks is an important objective in the context of precision agriculture, as it makes it easier for the farmer to plan ahead and efficiently use resources. Nevertheless, its implementation is labour-intensive and involves the manual counting of the fruit present in the trees. While colour (RGB) has been widely shown to be successful and arguably sufficient for yield estimation in orchards, hyperspectral imaging (HSI) shows promise for more nuanced tasks such as disease detection, cultivar classification and fruit maturity estimation. Therefore, it is important to ask how appropriate is HSI for the task of yield estimation, with a view to performing all of these tasks with just one sensor. This paper presents a novel mango yield estimation pipeline using ground based line-scan HSI acquired from an unmanned ground vehicle. Hyperspectral images were collected on a commercial mango orchard block in December 2017 and pre-processed for illumination compensation. After tree delimitation and mango pixel identification, an optimisation process was carried out to obtain the best models for fruit counting, using mango counts obtained by manually counting the fruit on-tree, and using state-of-the-art RGB techniques for yield estimation. Models were validated and tested on hundreds of trees, and subsequently mapped. In testing, determination coefficients reached values of up to 0.75 against field counts (predicting 18 trees) and 0.83 against RGB mango counts (predicting 216 trees). These results suggest that line-scan HSI can be used to accurately estimate yield in orchards, especially in scenarios in which this technology is already chosen for the determination of other traits.
关键词： Field robotics,Computer vision,Lidar,Hyperspectral,Fruit counting
更新于2025-09-23 15:23:52
-
[IEEE 2018 IEEE 3rd International Conference on Signal and Image Processing (ICSIP) - Shenzhen, China (2018.7.13-2018.7.15)] 2018 IEEE 3rd International Conference on Signal and Image Processing (ICSIP) - Vehicle Detection, Tracking and Counting
摘要： Traffic congestion and occlusions are major problems nowadays in metropolitan cities which leads to an ever growing traffic accidents. Therefore, the need of traffic flux management in order to avoid these congestions, unnecessary time wastage and tragic accidents is very important. Traffic regulation by optimizing timing of traffic control signals is one of the solutions for this purpose. This paper presents a low cost camera based algorithm in order to control traffic flow on a road. The algorithm is based on mainly three steps: vehicle detection, counting and tracking. Background subtraction is used to isolate vehicles from their background, Kalman filter is used to track the vehicles and Hungarian algorithm is exploited for association of labels to the tracked vehicles. This algorithm is implemented on both daytime and night time videos acquiered from CCTV camera and IR camera. Experimental results show the efficacy of the algorithm.
关键词： vehicle detection,Computer vision,tracking,counting
更新于2025-09-23 15:23:52
-
[IEEE 2018 Eighth International Conference on Image Processing Theory, Tools and Applications (IPTA) - Xi'an, China (2018.11.7-2018.11.10)] 2018 Eighth International Conference on Image Processing Theory, Tools and Applications (IPTA) - FACE - Face At Classroom Environment: Dataset and Exploration
摘要： The rapid development in face detection study has been greatly supported by the availability of large image datasets, which provide detailed annotations of faces on images. However, among a number of publicly accessible datasets, to our best knowledge, none of them are specifically created for academic applications. In this paper, we propose a systematic method in forming an image dataset tailored for classroom environment. We also made our dataset and its exploratory analyses publicly available. Studies in computer vision for academic application, such as an automated student attendance system, would benefit from our dataset.
关键词： image dataset,face recognition,face detection,computer vision,data collection,educational data mining,automated attendance system
更新于2025-09-23 15:22:29
-
[IEEE 2018 28th International Conference on Field Programmable Logic and Applications (FPL) - Dublin, Ireland (2018.8.27-2018.8.31)] 2018 28th International Conference on Field Programmable Logic and Applications (FPL) - Time-Shared Execution of Realtime Computer Vision Pipelines by Dynamic Partial Reconfiguration
摘要： This paper presents an FPGA runtime framework that demonstrates the feasibility of using dynamic partial reconfiguration (DPR) for time-sharing an FPGA by multiple realtime computer vision pipelines. The presented time-sharing runtime framework manages an FPGA fabric that can be round-robin time-shared by different pipelines at the time scale of individual frames. In this new use-case, the challenge is to achieve useful performance despite high reconfiguration time. The paper describes the basic runtime support as well as four optimizations necessary to achieve realtime performance given the limitations of DPR on today's FPGAs. The paper provides a characterization of a working runtime framework prototype on a Xilinx ZC706 development board. The paper also reports the performance of streaming vision pipelines when time-shared.
关键词： partial reconfiguration,realtime time-sharing,computer vision
更新于2025-09-23 15:22:29
-
Stereovision monitoring of deflection of concrete beam strengthened with ultraviolet-cured glass-fiber reinforced polymer in a destructive test
摘要： To monitor three-dimensional structural displacements in civil engineering, a stereovision displacement measurement method based on structure coordinate system is proposed in the present paper, and the absolute displacements of structure can be obtained through establishing the structure coordinate system and coordinate transformation. The center identification algorithm for circular target is studied to acquire the subpixel coordinates of center by combining Canny algorithm and Zernike algorithm. The epipolar constraint is introduced to conduct stereo matching of initial image pairs, and Kalman filtering and neighborhood searching algorithm are both employed to track circular targets on the left and right sequence images. To validate the effectiveness of the proposed method, a destructive test of concrete beams strengthened with ultraviolet-cured glass fiber reinforced polymer is performed in lab. Results show that the load–displacement curves obtained by the proposed stereovision method and linear variable differential transformer agree with each other; this verifies that the proposed stereovision method is feasible and effective for monitoring structural displacement in a destructive test.
关键词： Structural health monitoring,stereovision,computer vision,displacement measurement
更新于2025-09-23 15:22:29
-
Hardware implementation of digital image skeletonization algorithm using FPGA for computer vision applications
摘要： This paper proposed a method for digital image skeletonization of 2-D image of size 8 (cid:1) 8 and its implementation on Field Programmable Gate Array (FPGA). The time required to execute the proposed algorithm for 8 (cid:1) 8 dimension image on FPGA recon?gurable hardware is 4.815 ns, maximum output required time after clock: 4.075 ns, maximum frequency: 207.684 MHz, minimum input arrival time before clock: 2.284 ns. These values are for Vertex 5 FPGA board. This proposed algorithm ?nds applications in pattern recognition, computer vision, image matching and so on. This method can used in real time image processing applications. This algorithm may be extended for 3-D images and FPGA architecture may be proposed accordingly.
关键词： Computer vision,Gray scale images,2-D image,Skeleton,FPGA
更新于2025-09-23 15:22:29
-
Context-Aware Depth and Pose Estimation for Bronchoscopic Navigation
摘要： Endobronchial intervention is increasingly used as a minimally invasive means of lung intervention. Vision-based localization approaches are often sensitive to image artifacts in bronchoscopic videos. In this paper, a robust navigation system based on a context-aware depth recovery approach for monocular video images is presented. To handle the artifacts, a conditional generative adversarial learning framework is proposed for reliable depth recovery. The accuracy of depth estimation and camera localization is validated on an in vivo dataset. Both quantitative and qualitative results demonstrate that the depth recovered with the proposed method preserves better structural information of airway lumens in the presence of image artifacts, and the improved camera localization accuracy demonstrates its clinical potential for bronchoscopic navigation.
关键词： Computer Vision for Medical Robotics,Deep Learning in Robotics and Automation,Visual-Based Navigation,Visual Learning
更新于2025-09-23 15:22:29
-
Encyclopedia of Robotics || Vision for the Marine Environment
摘要： Vision for the Marine Environment refers to underwater imaging hardware and algorithms that enable the perception of the subsea environment for marine science applications, inspection and intervention. For a long time, optical cameras have been used in ROVs to provide the user with visual feedback of the operational scene. Conversely, AUVs have been traditionally equipped with sonar imaging systems, for two main reasons. First, the range of acoustic imaging is significantly higher, and second, as a consequence, they can work at a safer altitude, while the AUV follows the bottom profile. Nevertheless, during the last decades, vision systems have become smaller and more power-efficient, and the robot hardware has become more powerful and capable of storing the images onboard. Nowadays, commercial AUVs may be equipped with vision systems able to provide high-resolution seafloor imagery in clear waters. A single AUV survey may provide many thousands of images making it tedious to analyze the results. Image processing techniques may significantly improve the quality of the images and combine them into maps enabling faster interpretation. While the role of vision in commercial off-the-shelf robots is mostly passive (they gather data that is later processed after the mission), there have been research contributions aimed at incorporating computer vision algorithms to make the robot more autonomous, adapting its behavior online depending on the sensed situation. Algorithms for pipe/cable tracking, visual odometry, station-keeping, or target-tracking among others are paving the way toward an enhanced breed of autonomous robot, exploiting vision as a primary sensor modality.
关键词： Underwater computer vision,Underwater optical sensing and processing
更新于2025-09-23 15:22:29
-
[IEEE 2018 9th International Conference on Computing, Communication and Networking Technologies (ICCCNT) - Bangalore (2018.7.10-2018.7.12)] 2018 9th International Conference on Computing, Communication and Networking Technologies (ICCCNT) - A Perceptual Field of Vision, Using Image Processing
摘要： Currently, estimated facts state that there are more than 285 million visually impaired people around the globe, of which 39 million are blind and the others have a low vision [1]. Approximately 90% of people suffering from blindness are from low-income backgrounds [2]. The main aim of this paper is to provide an efficient visual platform to enhance the perception of the surroundings for a visually impaired user. This is achieved by using the concept of real-time image and video processing. This data is analyzed and compared along with the database which consists of pre-stored data that aids in recognition of the captured image. A head mount camera is used to capture an image on a real-time basis whenever desired. The camera is placed to provide a maximum field of vision and to eliminate the blind spot. The captured image is then processed and compared with the information stored in the database, providing an audio output indicating the desired information. Audio output is provided through bone conduction headphones which communicate audio signals directly with the inner ear. This keeps the outer ear free to be sensitive to the surroundings. A distress alert mechanism is also included as a safety measure to the blind at times of danger. It helps in sending messages containing distress alert signal which contains current location of the user.
关键词： virtual vision,computer vision,audio processing,distress alert,bone conduction,image processing,database
更新于2025-09-23 15:21:21