Design And Research Of The Quantitative Pouring System Based On Image Processing

High temperature of molten metal in the production of castings lead to the environment of pouring work very bad. Fortunately, automatic quantitative pouring can greatly improve the conditions of pouring. Meanwhile the pouring process directly affects the quality of castings. In the pouring process,insufficiency of liquid metal that poured into the casting cavity will produce defective castings, and over-pouring, leading to molten metal overflow from the pouring cup will not only causes the waste of metal, but also causes the damage of casting production line and conveyor belt, even causes the worker injury accident. As is known to all, common liquid level sensor can not be used to detect the height of the metal liquid level in the pouring cup, because of the high temperature of molten metal, so it is hard to realize full automation of pouring. At present, semi-automatic pouring with artificial participation(i.e. artificial quantitative) is widely used. Image processing technology provides powerful technical support for the automatic control of production process. In order to realize quantitative control of pouring process, we proposed the method of image processing for the automatic detection of the height of liquid metal level in the pouring cup because image processing technology can effectively solve the problem of sensor that used to detect liquid metal level. It has important practical significance and application value to achieve precise and quantitative automatic pouring.The design of automatic quantitative pouring systems consists of the hardware system and the software system. Hardware system is mainly composed of camera, image acquisition card, stepper motor, data input and output interface board and computer. The software system is designed by Delphi7, including interface design and writing program code. The whole system will complete theimage acquisition and processing, signal generation, pouring cup parameter setting, liquid level height detection, pouring equipments control and so on.In this paper, we analyze the existing methods of quantitative pouring, and then based on the combination of the non-contact image sensor and the computer image processing method, puts forward to a new detection method of molten metal liquid level in pouring cup. The principle of this system is useing a CCD image sensor to collect the image of molten metal surface in the pouring cup during the pouring process, based on the method of image processing to detect pixel value of each pixel point on the calibration track line and the coordinates of the current pixel point, and then calculate the current height of the liquid level and its growth speed according to the coordinates of the current pixel point. In order to achieve the purpose of automatic detection and control of the liquid level in the pouring cup, so as to realize automatic quantitative pouring of molten metal, then controller adjusts the motion of the stopper rod of the pouring ladle to control the opening degree of the ladle mouth based on the height of the liquid level and its growth speed in the pouring cup.To test the accuracy of liquid level detection of automatic quantitative pouring system. Firstly, tests was conducted with liquid at room temperature,then judge the sensitivity of the system for liquid level with detection by comparing actual liquid level and detected liquid level in the cup. The results show that error of the system is 1.12 %. On this foundation, experiments was carried out with molten aluminum alloy, the experimental results show that the error of the system is 2.97 %, less than 5 %, and is negative deviation, and the system error is in the allowable error range, so the system can satisfy the requirements in the production of casting by automatic quantitative pouring.