For concrete structure diagnostic inspection, ultrasonic velocity tomography method is generally used. This research was carried out for a new neural network tomography analysis method. This test was applied to a concrete test specimen having an interior defect portion by expandable foaming beads. In ultrasonic velocity tomography, many measurement lines are necessary. The new trial of the reduction of labor and the rationalization of the measurement is necessary too. In calculating the average ultrasonic velocity of each concrete inspection block, there is still plenty of room for improvement of the interior diagnostic probes. By using a neural network, this research worked to be a progressive method by having an inference function that was estimated from the tendency of the observation result. Furthermore, a neural network is recommended to evaluate the results from fusion processing with the internal ultrasonic velocity and surface thermography index. And the different nondestructive inspection results are able to fuse by the neural network. Additionally, there is a necessity for an inferential nondestructive combination evaluation method. Lastly, a new inference fusion method is significant technical skill in the concrete quality grasp from the surface of thermography and ultrasonic velocity for visualizing concrete interior.

Related Publications