| Author: | Wang, Shuai |
| Title: | Towards eye-in-hand continuum robotics : concurrent measurement and control in constrained environments |
| Advisors: | Chu, K. Henry (ME) |
| Degree: | M.Sc. |
| Year: | 2023 |
| Subject: | Endoscopy Robotics in medicine Hong Kong Polytechnic University -- Dissertations |
| Department: | Department of Mechanical Engineering |
| Pages: | xvi, 45 pages : color illustrations |
| Language: | English |
| Abstract: | Medical care has been the top priority of social civilization in all parts of the world. With the advancement of science and the development of robot technology in recent years, it is a future development trend to gradually apply robot technology to the medical field. In particular, robot systems can assist doctors in judgment and treatment, and improve medical efficiency and accuracy. In the future, robot-assisted medical systems will also be an indispensable link in future medical care. Endoscopes have also gone through various stages of traditional rigid endoscopes, flexible passive bending endoscopes, and flexible active bending endoscopes, which integrates traditional optics, ergonomics, precision machinery, modern electronics, mathematics, and software. Compared with traditional endoscopes, controllable flexible endoscopes can reduce secondary trauma to the human body. It can be used to see lesions that cannot be seen on X-rays, it is very useful for doctors combining the kinematics of the endoscopic robot with the data collected by the camera can also give doctors a more intuitive feeling about the lesion. The first endoscopes were made of rigid tubes and were invented more than 100 years ago. Although they have gradually improved, they are still not widely used. Later, in the 1950s, endoscopes were made of flexible tubes that could easily bend around the corners of the body. In 1965, Harold Hopkins installed a cylindrical lens on the endoscope to make the field of view clearer. Today’s endoscopes usually have two fiberglass tubes, and light enters the body through one of them. The doctor’s observation is made through another tube or acquired from a video camera, and some endoscopes even have tiny integrated circuit sensors that feed back what they see to a computer. At present, endoscopes can be roughly divided into two categories according to whether their bodies can be bent: bendable and non-bendable according to its structure. In recent years, non-bendable endoscopes have been gradually replaced by bendable endoscopes in applications because bendable endoscopes can pass through the natural orifice of the human body, minimizing the secondary injury of patients during treatment. Besides the flexible structure, the visual information obtained by the endoscope is also a treasure in the whole mission. However, the information seen directly from 2D images is not enough, it is difficult for the doctor to have an intuitive feeling for information such as the size of the lesion. Therefore, we use the controllability of the bendable endoscope, cooperated with the image information obtained by the camera in the endoscope, and using the optimization algorithm to estimate the depth information of the features relate to the camera. In this article, a single-joint continuum robotic endoscope with two degrees of freedom, which use constant curvature model to establish its kinematics, combing with a resolution of 400*400 camera placed at the end of the endoscope. Successive key frames are collected from monocular and constraint equations for feature depth estimation are established to find information that the position of targets in the camera coordinate system. Based on principles that: 1)The size of features observed at different camera positions remains the same. 2)And a point in feature is invariant relative to the position of the world coordinate. Constraint equations can be built and estimating the position of the feature point relative to the camera. When building the kinematic model, a stiffness factor K is introduced to describe the bending effect of different materials in various robots. In the first experiment, I find stiffness factor K related to the material in endoscope. NDI Aurora sensor was used to collect multiple sets of moving data. By using the truth that the calculated displacement relate to camera was the same as that of monocular driven by endoscope, and the coefficient value K was calculated with the least square method. In the second experiment, a screen displaying features (four different color points) is placed at eight positions surrounding the endoscope. Camera collects 4 frames of images and calculates the points depth data. For each position that the pattern placed, depth estimation experiment is repeated five times to obtain feature information and compare it with the true value measured from sensors. Due to it is really hard to locate the feature’s position relate to world coordinate. I use estimated depth information that the feature position related to the camera coordinate calculate the distance between opposite feature points compared with the real value of the length of the line segment formed by the feature points to validate accuracy. The average error of the distance between features is about 0.3mm, and the estimated value is volatile. By optimizing the feature recognition algorithm and improving the pixel resolution of the endoscope, the disturbance noise can be reduced and the estimation accuracy can be improved. In the end, an open-loop trajectory motion was tested with a positional accuracy within 3 mm. |
| Rights: | All rights reserved |
| Access: | restricted access |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| 7477.pdf | For All Users (off-campus access for PolyU Staff & Students only) | 6.29 MB | Adobe PDF | View/Open |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/13009