oversampling

In modern conditions, the problem of mine detection remains one of the most urgent due to the serious threat to the life and health of people in contaminated areas. This paper presents an approach to mine detection using a hybrid neural network CNN+BiLSTM+Attention, which analyzes B-scan signals received from ground-penetrating radar systems. To improve the quality of training with a limited amount of data, image augmentation was used, which includes shifting, reflecting, scaling, and adding noise.

 A new method for rapid automatic detection values of crossing-over operations in the of the image preprocessing tasks, which based on the divergence matrix operators. Experimental results shown high resistance of method to image processing with fluctuation of intensity function. Comparing the results of the proposed method with the results for the existing, showed acceleration automatically select the crossover factor that significantly reduces the necessary computing power for its operation. This leads to the possibility of he effective usage in preprocessing of large-size images.