The Monitoring System Of Ultra-precision Machining Environment Based On Image Recognition

Advanced optical components such as high-precision flat and aspheric optical lens, are widely used in high-tech fields of defense and civilian. And as the typical brittle and hard materials, the ultra-precision machining environment must be strictly controlled to achieve the stable machining accuracy and surface quality. However, the auxiliary equipments of machine tool generally produce heat, vibration and so on, which affect the ultra-precision machining process. As an example, a large size and high precision surface grinder named2MK1760was developed independently by our team, and the heat and vibration caused by the auxiliary equipments include precise oil cooling machine, oil pump and precision water-cooled machine, seriously affect the process of ultra precision machining. Therefore, the auxiliary equipments of ultra-precision machine tools are always isolated, such as placed outside the clean workshop and sealed in soundproof boxes. As a result, it is very difficult to monitor ultra-precision machining environment and detect the emergent problems. For this purpsoe, it is very essential to design an efficient and steady monitoring system of ultra-precision machining environment.Because auxiliary equipments of ultra-precision machine tools generally have no communication interface to computer, the monitoring system is designed in this paper to obtain the ultra precision machining environment data through the real-time shooting machine tool equipment instrument image, and processing and recognition of image. The monitoring system consists of upper computer and lower computer, and the lower computer whose processor is TMS320DM642mainly deals with image processing and recognition, the upper computer mainly executes data analysis and processing. Image processing and recognition of dial instrument and digital instrument is the key technology of monitoring system. So the recognition algorithm of digital images is in-depth studied, and then a window scanning algorithm is presented, which can detect the indicator of meters rapidly. Besides, the programs in upper computer and lower computer are designed, including image processing and recognition. Finally, the experiments for recognitions of dial instrument and digital instrument are carried out on the monitoring system, and the result well reflects that both the image recognition algorithm and the monitoring system are efficient and steady.