This article proposes a new technique for determining the rate of contamination. First, a generative adversarial neural network (ANN) parallel processing technique is constructed and trained using real and secret images. Then, after the model is stabilized, the real image is passed to the generator. Finally, the generator creates an image that is visually similar to the secret image, thus achieving the same effect as the secret image transmission. Experimental results show that this technique has a good effect on the security of secret information transmission and increases the capacity of information hiding. The metric signal of noise, a structural similarity index measure, was used to determine the success of colour image-hiding techniques within ANN. The results of the ANN were in sequence: 41.2813, 0.6914. The results of the ANN were in sequence 41.2813, 0.6914. These results provide insights into how well the hidden information is concealed within the image and the extent to which the visual integrity of the image is preserved.
Design an Efficient Neural Network to Determine the Rate of Contamination in the Tigris River in Baghdad City
Artificial Neural Networks
Physical Attributes
Unconstrained Optimization
Levenberg- Marquardt (LM) Algorithm
Contamination