Edge-guided refined saliency object segmentation method and system and equipment

Abstract

The invention relates to the field of image processing, and particularly to an edge-guided refined saliency object segmentation method and system and equipment, and aims to solve the problem that segmentation results in the prior art are relatively rough. The saliency object segmentation method of the invention comprises: obtaining different-scale segmentation mask features of an input image through forward propagation of a segmentation mask sub-network; obtaining different-scale edge detection features of the input image through forward propagation of an edge detection sub-network; and fusingthe segmentation mask features and the edge detection features through a backward fusion branch network to obtain a saliency object segmentation result and an edge detection result. In addition, a focus cross-entropy loss function is utilized to supervise a training process when the segmentation mask sub-network is trained, and the sub-network is enabled to focus on error-prone classification samples of object edge regions and the like. The saliency object segmentation result of the method is more refined, and retains more accurate edge region detail information.

Description

technical field [0001] The invention relates to the field of image processing, in particular to an edge-guided refined salient object segmentation method, system, and device. Background technique [0002] Traditional salient object segmentation methods rely on hand-designed features, but hand-designed features are difficult to represent complex image changes: such as deformation, occlusion, lighting condition changes, complex background effects, etc. Thanks to the rapid development of deep convolutional neural networks, the performance of image salient object segmentation methods based on fully convolutional neural networks has also been significantly improved. However, due to the inherent limitations based on the full convolutional neural network, the segmentation results of image salient objects are often rough, especially in the object edge area, and the segmentation results are difficult to preserve good detail information in the object edge area. [0003] The fully con...

AI Technical Summary

Problems solved by technology

However, due to the inherent limitations based on the full convolutional neural network, the image salient object segmentation results are often rough, especially in the object edge area, and the segmentation results are difficult to preserve good detail information in the object edge area.

[0003] The fully convolutional neural network is composed of a series of fully convolutional layers and downsampling layers stacked on top of each other. While these downsampling layers aggregate semantic information, they also greatly reduce the original resolution and lose most of the spatial positioning. Information, although the last upsampling layer or deconvolution layer, it is difficult to accurately restore the lost spatial detail information, which is extremely unfavorable for image pixel-level classification tasks such as salient object segmentation that require accurate spatial position information, and the segmentation results Relatively rough, it is difficult to preserve spatial details, especially the edge area of ​​​​the object

Images