Chao Hu

Due to the problem of performance constraints of unsupervised video object detection, its large-scale application is limited. In response to this pain point, we propose another excellent method to solve this problematic point. By incorporating motion characterization in unsupervised video object detection, detection accuracy is improved while reducing the computational amount of the network. The whole network structure consists of dual-stream network, motion guidance module, and multi-scale progressive fusion module. The appearance and motion representations of the detection target are obtained through a dual-stream network. Then, the semantic features of the motion representation are obtained through the local attention mechanism in the motion guidance module to obtain the high-level semantic features of the appearance representation. The multi-scale progressive fusion module then fuses the features of different deep semantic features in the dual-stream network further to improve the detection effect of the overall network. We have conducted numerous experiments on the three datasets of DAVIS 16, FBMS, and ViSal. The verification results show that the proposed method achieves superior accuracy and performance and proves the superiority and robustness of the algorithm.

The intention of leveraging Radio-Frequency (RF) resources for diverse sensing purposes has grown increasingly keen, thanks to the ever-expanding deployment of IoT devices using RF communications to maintain connectivity. Whereas the original idea was to integrate sensing and communications (a.k.a. ISAC) for individual IoT devices, most proposals in the past decade simply converted such devices (e.g., Wi-Fi or LoRa) into RF sensors without attending to their communication nature. Therefore, we argue that the device architecture has to be overhauled so as to achieve the ISAC ambition. To this end, we propose ISACoT as the framework for enabling ISAC over IoT devices. We categorize ISACoT's extensions over existing devices into four aspects, namely time, frequency, space, and protocol. We argue that, as the multistatic communication infrastructure of IoT is adverse to device-free sensing (e.g., lack of precise time synchronization), the keystone of ISACoT should be operating sensing in a monostatic mode (like radar).We tackle the fundamental time aspect based on Wi-Fi first, but leave the other three as challenges for exploration. For each of the three aspects, we present our proposals along with verification experiments, while putting forward challenges along with potential solutions. In particular, the protocol aspect discusses how ISACoT can be generalized to other IoT protocols beyond Wi-Fi.

Electromagnetic tracking systems (EMTSs) are widely employed in medical instruments, virtual reality, game control, etc. An EMTS mainly consists of three parts:magnetic field generator connected with transmitting coils, control unit, and sensing and computing unit. In general, the transmitting coils and the induction coils of an EMTS are connected to the same control unit to determine an induced electromotive force resulted from transmitting coil. In this article, a novel EMTS is presented, whose induction coils are separated from the control unit; hence, the wireless pose tracking can be achieved. To improve the tracking speed and determine an induced electromotive force resulted from transmitting coil, the transmitting coils are simultaneously excited by nine-channel sinusoidal signals with different frequencies, instead of the switching channels in the chronological order. Thus, five-degree-of-freedom tracking of the miniature induction coil is implemented. The experimental results show that the average position and the orientation errors are less than 2.3 mm and 0.2°, respectively, which indicates that the proposed EMTS has potential prospects for both medical and industrial applications, especially for the magnetically controlled capsule robot.

Magnetic tracking approach (MTA) is mainly based on the observation of the magnetic field produced by a magnetic tracking target. A permanent magnet is usually employed as the tracking target, and magnetic dipole model is used to estimate the magnetic field. In this study, we introduce an intuitive method to solve the magnetic tracking problem based on graph optimization. A graph is constructed to formulate the tracking problem, whose nodes correspond to the poses of the permanent magnet at different time and whose edges represent constraints derived by sensor measurements. The accuracy and robustness of MTA play a vital role in biomedical and industrial applications. However, a potential abnormal edge will appear in the graph when a magnetic sensor is disturbed by an interference source like a small magnet, which will deteriorate the localization accuracy of MTA. To improve its robustness, Huber cost function is used to reduce the weight of the abnormal edge. Finally, experiments were carried out to verify the performance of the proposed approach with and without magnetic field interference. Comparing the localization errors with the MTA based on the standard Levenberg-Marquardt algorithm, the results illustrate that the proposed approach could achieve superior localization accuracy and anti-interference ability.

Recently, a variety of positioning and tracking methods have been proposed for capsule robots moving in the gastrointestinal (GI) tract to provide real-time unobstructed spatial pose results. However, the current absolute position-based result cannot match the GI structure due to its unstructured environment. To overcome this disadvantage and provide a proper position description method to match the GI tract, we here present a relative position estimation method for tracking the capsule robot, which uses the moving distance of the robot along the GI tract to indicate the position result. The procedure of the proposed method is as follows: firstly, the absolute position results of the capsule robot are obtained with the magnetic tracking method; then, the moving status of the robot along the GI tract is determined according to the moving direction; and finally, the movement trajectory of the capsule robot is fitted with the Bézier curve, where the moving distance can then be evaluated using the integral method. Compared to state-of-the-art capsule tracking methods, the proposed method can directly help to guide medical instruments by providing physicians the insertion distance in patients’ bodies, which cannot be done based on absolute position results. Moreover, as relative distance information was used, no reference tracking objects needed to be mounted onto the human body. The experimental results prove that the proposed method achieves a good distance estimation of the capsule robot moving in the simulation platform.

Pose tracking based on a permanent magnet is not limited by line-of-sight. In general, a high-remanence and axially magnetized cylindrical permanent magnet is utilized as the magnetic source, which is regarded as a magnetic dipole model. However, this magnetic tracking approach cannot estimate the rotation information along the direction of the magnetic moment. Thus, only three-dimensional (3D) linear motion and 2D rotational motion of the magnet can be tracked. On the other hand, inertial sensors provide effective 3D orientation tracking. In this study, an inertial measurement unit was attached on the surface of a permanent magnet. Then the tracking information of these two modes was fused based on the adaptive weighted fusion algorithm. Experiments show that the full 6D pose is acquired with an improved orientation accuracy, where the average orientation errors for α, β, and γ, obtained by the fusion algorithm, were 0.81°, 0.93°, and 1.26°, respectively. This study is helpful to those who want to accurately track the full 6D pose of an object.

Predictive Function Control of the Single-Link Manipulator with Flexible Joint | InTechOpen, Published on: 2012-06-13. Authors: Zhihuan Zhang and Chao Hu

A capsule endoscope robot is a miniature medical instrument for inspection of gastrointestinal tract. In this paper, we present image compact representation and preliminary panoramic representation methods for the capsule endoscope. First, the characteristics of the capsule endoscopic images are investigated and different coordinate representations of the circular image are discussed. Secondly, effective compact representation methods including special DPCM and wavelet compression techniques are applied to the endoscopic images to get high compression ratio and signal to noise ratio. Then, a preliminary approach to panoramic representation of endoscopic images is presented.