| Aluminum foil plays an important role in daily life,the defects of which not only affect the appearance of the aluminum foil product,but also affect its performance.Defects detection on the surface of the aluminum foil can get the production quality of the aluminum foil coil evaluated,and the defective products can be picked and eliminated to improve the production efficiency.In this paper,for the detection requirements of automation of aluminum foil detection,a detection system capable of effectively detecting aluminum foil defects is designed,which has been employed in practical industrial detection.In this paper,a light field imaging system with good illumination effect is designed for the optical properties of aluminum foil with strong reflection and non-transmission.In the bright field or dark field conditions,the contrast between defects and background are not the same.By setting up two sets of light sources at the same inspection station,the bright field and the dark field are respectively formed to achieve better effect of various defects.To acquire accurate defect information,the median filtering algorithm with better filtering effect is selected,and the edge of the image is detected by the Otsu algorithm.Based on the obtained defect morphology information,the detected defects are classified.According to the classification results,image detection algorithm are designed for different defects.Finally we calculate the defect information on the aluminum foil roll based on the detection result of the aluminum foil.Aiming at the need of human-computer interaction in the system,the software interaction interface was designed.According to the technical indicators of the detection,the hardware of the detection system was designed.Finally,the comparison between results from the system and worker verify the detection accuracy of the system. |
