Image processing method and device

Abstract

The invention provides an image processing method and device. The method comprises steps of cquiring the air transmissivity corresponding to the target object according to the depth information of the target object in an original image, a random noise generation algorithm and a preset air scattering parameter; and generating a fog image according to the air transmittance, the air illumination parameters in the original image and the original image. According to the method and the device, the fog image can be automatically generated according to the original image, so that the data volume of the training data used for training the model can be increased, and fog in the fog image generated by combining the depth information of the target object and the random noise generation algorithm conforms to a real scene with fog in reality; for example, the fog in the fog image is non-uniformly distributed, and the fog distribution is more random and natural, so that the generalization ability of the trained model in a real scene with fog in reality can be improved. And manual participation is not needed when the fog image is generated according to the original image, so that the manual workload can be reduced, and the labor cost is reduced.

Description

technical field [0001] The present application relates to the field of computer technology, in particular to an image processing method and device. Background technique [0002] Machine learning and other artificial intelligence algorithms based on neural network models have been widely used in related fields such as video images, speech recognition, and natural language processing because of their strong fitting capabilities and end-to-end global optimization capabilities. [0003] However, the generalization ability of the neural network model depends on the amount of training data used to train the neural network model. The more training data, the higher the generalization ability of the trained neural network model. [0004] In this way, before training the neural network model, a large amount of training data needs to be prepared. However, in the prior art, it is currently necessary to manually collect training data and mark the training data, but the collection and ma...

AI Technical Summary

Problems solved by technology

[0004] In this way, before training the neural network model, a large amount of training data needs to be prepared. However, in the prior art, it is currently necessary to manually collect training data and mark the training data, but the collection and marking of a large amount of training data requires It consumes a very large amount of manual work, resulting in high labor costs

Images