Design Of The Tube Sampling Robot Based On Embedded Video System And Research On Video Denoising Algorithm

The central air-conditioning system is actually becoming the lung of modern buildings by controlling their air metabolism. As people are more and more concerned about the environment they live in, the indoor air pollution produced by the ventilating system of the central air conditioner draws more and more people's attention and puts forward higher requirements on the hygiene supervision and management departments. The development of a quantitative sampling robotic system for ventilation tube offers an accurate and convenient detection method for the departments, and also fills the gaps in research and application area both at home and abroad.According to the provisions of the Ministry of Health and the requirements analysis, a robotic system, which consists of a robot, a computer and a control box, is designed for working in ventilation tube, conducting quantitative aseptic sampling. The robot is driven by two DC motors. Two cameras and a sampling system are installed on it. By operating the control box, the operator can drive the remote mobile robot to the designated place of the rube and collect dust on the floor, while the cameras transmit real-time video to the interaction system on the computer.In order to get clear pictures in the tube with low illumination, a new denoising algorithm based on neuron model is proposed in this thesis. With the maximum probability theory from neuron model applied in the algorithm, the random noise in a slow-motion video can be significantly reduced. In this thesis, the validity of the denoising algorithm is proved by applying it on the standard gray image in which man-made random noise is added. The algorithm is also valid on RGB images, which expands its scope of application.Compared with the traditional method, the robotic system described in this thesis reduces the complexity and environmental interference for dust sampling in tube. It is flexible, powerful, accurate and easy to control, the image gained is very clear. Results from experiments in lab prove the effectivity of the system and can satisfy the requirements of criteria and fulfill the needs of industrial tasks.