Double-hemispheric capsule robot attitude detection method based on magnetic vector error calibration image

Abstract

The invention discloses a double-hemispheric capsule robot attitude detection method based on a magnetic vector error calibration image and belongs to the technical field of automation engineering. The method uses the error distribution characteristic of a spatial universal rotation magnetic vector, firstly, the least azimuth error direction of the magnetic vector is determined as an initial imagecalibration position of the attitude of a double-hemispheric capsule robot, image information before and after attitude adjustment is acquired by a camera, recognition and matching of feature pointsof the image are performed, the relative attitude change before and after attitude adjustment is obtained by combination of epipolar geometry and a rotation matrix, and the actual attitude of the double-hemispheric capsule robot is effectively estimated. The method can estimate the attitude of the double-hemispheric capsule robot in wide gastrointestinal environments only by the camera carried bythe robot without any external hardware resources.

Description

Technical field [0001] The invention belongs to the technical field of automation engineering, and relates to an initial image calibration position of a double-hemisphere capsule robot with a minimum error orientation determined by a space universal rotating magnetic vector error model, and the initial calibration and posture adjustment obtained by an embedded wireless camera The feature point recognition and matching of the two image information, combined with the epipolar geometric constraints, obtains the posture detection method of the axis orientation of the double hemisphere capsule robot in the ample environment such as the gastrointestinal tract and colon. Background technique [0002] In the field of gastrointestinal endoscopy detection, due to the limitations of traditional endoscopic detection methods, it cannot detect areas such as the small intestine with complex physiological environments, and the detection process will bring great pain to the patient, and repeated u...

AI Technical Summary

Problems solved by technology

[0003] The working environment of the capsule robot is the human gastrointestinal tract with complex physiological structure. At present, the Pillcam series capsule robot produced by Israel Given Imaging Company and the OMOM capsule robot produced by Chongqing Jinshan Company have been put into clinical application. The position and attitude of the passive capsule robot are uncontrollable, and the detection time is up to 7-8 hours

[0004] The above-mentioned passive capsule robot has the following problems: (1) the detection time is long

It needs to move forward with the help of intestinal peristalsis, and cannot quickly reach the corresponding lesion area, and it is easy to miss the lesion area, resulting in missed detection

(2) It takes up a lot of image memory resources

However, the posture information obtained by vision is mainly the movement of the capsule robot in a narrow environment such as the intestinal tract, which are the rotation around its own axis and the displacement along the axis of the intestinal tract, respectively, and it cannot obtain the capsule robot’s position in the ample environment. More complex attitude information such as pitch angle and roll angle has certain limitations, and it also poses a great obstacle to identifying lesion areas

[0013] At present, no one has proposed to use the uniformity of the universal rotating magnetic field and the follow-up of the dual-hemispheric capsule robot in the magnetic field. Under the condition that there is no need to install any sensors or other devices in the dual-hemispheric capsule robot, the magnetic vector error model and A detection method for determining complex attitude information such as pitch angle and roll angle of a dual-hemispheric capsule robot in an ample environment using image features acquired by an embedded camera

Images