Image instance segmentation method and device, electronic equipment and storage medium

Abstract

The invention relates to an image instance segmentation method and device, electronic equipment and a storage medium, and relates to the technical field of computer vision. The image instance segmentation method is used for solving the problem of inaccurate instance segmentation in related technologies. The image instance segmentation method of the present disclosure includes the steps: inputtingthe feature map of the target into a trained segmentation network; based on the segmentation network, determining a sub-image of a target in the feature map, wherein the segmentation network is trained in the following manner; carrying out down-sampling on the sample graph; processing the down-sampled sample graph based on the hole convolution kernels with different expansion rates; determining sample graphs of different receptive fields, fusing the sample graphs of the different receptive fields to obtain a fused sample graph, performing up-sampling and instance segmentation on the fused sample graph to determine a first sub-image, and training an initial segmentation network according to the first sub-image and the marked second sub-image; and carrying out position marking of the sub-images in the image, and because pixel information of pixel points in the feature maps of different receptive fields is large, instance segmentation is more accurate.

Description

technical field [0001] The present disclosure relates to the technical field of computer vision, and in particular to an image instance segmentation method, device, electronic equipment and storage medium. Background technique [0002] Instance segmentation is an important part of computer vision, and many research works have been carried out to improve the accuracy of segmentation. The earliest instance segmentation method is a bottom-up method, such as the Deepmask method, which first predicts the segmentation candidate area, and then uses the classification network to classify. However, in this type of method, the segmentation process precedes the classification process, resulting in a slow instance segmentation process. , the accuracy is not high. The next FCIS (Fully Convolutional Instance-aware Semantic Segmentation) scheme combines segmentation candidate regions and object detection systems, which can improve the speed of instance segmentation, but such methods do no...

AI Technical Summary

Problems solved by technology

[0005] The present disclosure provides an image instance segmentation method, device, electronic equipment, and storage medium to solve the problem of inaccurate edge segmentation of instances in the instance segmentation process in the related art

Images