A new technology for detecting a tiny residual core in the small inner cavity of complex castings is proposed. The residual core is identified by using image recognition technology. Tracer processing and image signal ...A new technology for detecting a tiny residual core in the small inner cavity of complex castings is proposed. The residual core is identified by using image recognition technology. Tracer processing and image signal processing are combined to enhance the image contrast. The relationships between the concentration of tracer, the size of the residual core, the wall thickness of the castings and the contrast were obtained. Based on the experimental data, the minimum detectable amount of residual core under different conditions was obtained. The results show that the minimum detectable amount decreases from 4.398 mg to 0.438 mg for the 1.0 mm wall thickness casting when the concentration of tracer increases from 0% to 20%. The signal-to-noise ratio(SNR) of the detection results increases by 27.010 by means of average filtering and linear point operation. The subtraction of image and image background was performed, and then the boundary extraction was carried out to obtain a clear and reliable result. The experimental results show that the non-traced residual core cannot be detected for a blade with a thickness less than 5 mm. The residual core of 1 mm thickness can be barely identified by artificial recognition after tracer processing and image processing, while the residual core of 0.6 mm thickness can be detected clearly using image recognition technology.展开更多
基金supported by the National Natural Science Foundation of China(No.51475120)Major Program of National Natural Science Foundation of China(No.U1537201)
文摘A new technology for detecting a tiny residual core in the small inner cavity of complex castings is proposed. The residual core is identified by using image recognition technology. Tracer processing and image signal processing are combined to enhance the image contrast. The relationships between the concentration of tracer, the size of the residual core, the wall thickness of the castings and the contrast were obtained. Based on the experimental data, the minimum detectable amount of residual core under different conditions was obtained. The results show that the minimum detectable amount decreases from 4.398 mg to 0.438 mg for the 1.0 mm wall thickness casting when the concentration of tracer increases from 0% to 20%. The signal-to-noise ratio(SNR) of the detection results increases by 27.010 by means of average filtering and linear point operation. The subtraction of image and image background was performed, and then the boundary extraction was carried out to obtain a clear and reliable result. The experimental results show that the non-traced residual core cannot be detected for a blade with a thickness less than 5 mm. The residual core of 1 mm thickness can be barely identified by artificial recognition after tracer processing and image processing, while the residual core of 0.6 mm thickness can be detected clearly using image recognition technology.
