Neural network design
| Hi-index | 0.00 |
An effective NDT (Non-Destructive Testing) image analysis technique for detecting materials damage and defects existence has been developed successfully and applied to PVT images of composite structures. The developed technique is based on converting an image to its equivalent pixel values and then applying Search Through (ST) algorithm to the converted image such that the presence of damage in the composite structure and its extent can be easily verified. The technique has a novel approach to data analysis by employing intensity, RGB signal re-mix and wavelength variation of a thermally generated IR-beam onto the specimen under test which can be sensed and displayed on a computer screen as an image. Specimen inspection and data analysis are carried out through pixel level re-ordering and shelving techniques within a transformed image file using a sequence grouping and regrouping software system, which is specifically developed for this work.