Research On Detection Of The End Point Of Flame In Waste Incineration Based On Machine Vision

At present, the detection of end points of flame mainly depends on the view of human eyes through the holes to observe the fire in waste incineration system. This way to control the action of grate, is not only labor-intensive, high-risk, but also if adjusted furnace improperly, it will turn off even cause major security issues. so it is essential to detect the state of flame in waste incineration. Nowadays, domestic manufacturers are detecting the edge of flame artificially with a hole in incinerator, which is of great labor intensity, low efficiency and inaccuracy. In order to improve the current situation, it is necessary to establish a system of identifying the points of flame based on machine vision.Firstly, the process of waste incineration has been studied; compared with several ways of waste incineration methods, the way of pyrolysis is selected in this paper; According to the complexity of the garbage composition , the detection flame end points of waste incineration system based on machine vision has been established. The choice of camera, image transmission fiber and the style of image collection card have been confirmed; and industrial PC has been selected. Secondly, the way of how to detect end points of waste incineration flame based on machine vision is studied systematically. compared with several ways of filtering methods, median filtering is used to strengthen the images of waste incineration with the consideration of the system environment; and then use segmentation based on histogram-based DFC to segment the image under the consideration of particularity of the images of flame of waste incineration; the noise existing in the image after separating used morphology method to de-noise and enhance the character of the image; Aiming at the requirements for measurement, the image edge detection algorithm based on phase information is used to extract the flame edge outline through the specific analysis and comparison of various edge detection algorithm.Finally, simple and applied operation interface have been designed according to the design principles and requirements of software system, by using VC++ ,and the software program for measurement based on image processing has been written. The proof shows that this detect system is easy to operate and has high application value.