Object Searching Of Wards Inspection Robot

With the help of smart space, using service robots in hospital wards inspection, which can greatly reduce the workload and mental pressure of nurses. It can also be able to notice the problems of the hospitalization, and the results could be realisticly significant. A robot ward inspection system of intelligent space in the hospital environment is proposed in this thesis. This system uses a variety of image segment methods to achieve the goal of specific objects and body parts searching in wards. The main contributions of this thesis are shown as follows:Firstly, design robot ward inspetion system:hardware support and robot navigation, software design of wards inspection, interactions between wards inspection system and server system, design of inspectioin process, camera calibration and targeting, and those established the basis for image segment algorithms addition.The key contents of this paper is specific object search algorithms in ward inspection, aiming at important objects such as infusion device, body dectection, medical instrument. It is roughly divided into three parts:the infusion device segmentation algorithm basing on the color model, the inspection tasks basing on skin detection, the medical instrument segmentation algorithm basing on the geometric characteristics. In the color mode research, this paper analyzes color space mode, color histogram mode, color lookup table mode, finally lookup table mode is chosen to locate artificial markers, and estimates the position of infusion device. The inspection tasks basing on skin detection includes face-detection, medical needle area detection, detection of falling off bed, after the research of commonly used algorithms, an adptive skin detection algorithm basing on color space mode and Adaboost face-detection is desgined. The design of medical instruments segmentation algorithm is base on extraction of geometric characteristics. Through the research and comparison of commonly used image-segment algorithms, we designed different algorithms according to the specific environment and object characteristics. The accuracy and stability of these algorithms are proved by a large number of experiments, and the robot can complete the task of wards inspection successfully.