Optimization of single filter network on visual corrosion defect

Due to the challenging environmental conditions and characteristics, the complexity of the corrosion inspection operation increases. By using software image filter to enhance image data, the object recognition technique will be able to analyze the image data accurately. A selected software filter, wavelet de-noising has been identified to enhance image data for visual corrosion inspection application. Therefore, in order to obtain a better image enhancement, neural network is used for validation. The experiment result shows neural network wavelet de-noising filter used to enhance image in order to obtain better image for visual corrosion inspection is achievable, and gives desirable result in terms of Mean Square Error and Peak Signal to Noise Ratio. This project is focusing on corrosion inspection using image.