Hollow recognition method and device based on neural network, storage medium and terminal

Abstract

The invention discloses a hollow recognition method and device based on a neural network, a storage medium and a terminal, and the method comprises the steps: providing a training set and a test set,wherein the training set comprises a plurality of hollow training pictures containing hollows and a plurality of hollow-free training pictures not containing hollows, aand the test set comprises a plurality of test pictures; segmenting each hollowed-out training picture into a plurality of hollowed-out picture blocks, and segmenting each non-hollowed-out training picture into a plurality of non-hollowed-out picture blocks; training a neural network model based on the hollowed-out picture blocks and the non-hollowed-out picture blocks; performing a hollow recognition test on the test picture byusing the neural network model, wherein the neural network model comprises a convolution layer, a pooling layer, an activation layer and a 1 * N output layer. According to the invention, the identification accuracy and efficiency can be improved, and the identification cost is reduced.

Description

technical field [0001] The present invention relates to the field of image processing, in particular to a neural network-based hollowing out recognition method and device, a storage medium, and a terminal. Background technique [0002] At present, mainstream mobile phones have a portrait mode that simulates background blur. The core step is to obtain depth information through the main / sub-camera, and realize the gradient of blur through the depth of field, and finally achieve the effect of DSLR camera optical background blur. However, due to the complexity of the scene and the limitation of the current dual-camera technology, the hollowing problem often occurs in portrait blur photos, that is, the background near the edge of the human body should be blurred but it does not take effect. This situation will greatly affect the image. The quality of the effect degrades the user experience. [0003] In the current existing detection scheme, testers are required to manually evalu...

AI Technical Summary

Problems solved by technology

However, due to the complexity of the scene and the limitation of the current dual-camera technology, the hollowing problem often occurs in portrait blur photos, that is, the background near the edge of the human body should be blurred but it does not take effect. This situation will greatly affect the image. Quality of effect, degraded user experience

[0003] In the current existing detection scheme, testers are required to manually evaluate the severity of hollowing out according to the test scenario. This method has the problems of high cost and low accuracy, and it is difficult to meet the demand

Images