基于改进BiSeNet的语义分割算法

Semantic segmentation algorithm based on improved BiSeNet

语义分割在无人驾驶环境感知方面有重要的研究意义,需要同时兼顾语义和空间信息。针对提取语义信息时空间信息丢失不可恢复的问题,提出一种基于改进BiSeNet的语义分割算法。首先,针对空间路径信息冗余的问题,将输入图像转换为灰度图,并通过Sobel算子提取梯度信息后,采用卷积结构提取空间信息;然后,对两分支不同分辨率特征图进行逐步融合上采样。在Cityscapes数据集上对算法进行验证,分割精度达到65.79%,相比BiSeNet提高了3.11%。

Semantic segmentation has important research significance in perception of unmanned driving environment,and it needs to take into account both semantic and spatial information.Therefore,aiming at the problem of unrecoverable loss of spatial information when extracting semantic information,a semantic segmentation algorithm based on improved BiSeNet is proposed.Firstly,aiming at the problem of spatial path information redundancy,the input image is converted into grayscale image,and the gradient information is extracted through the Sobel operator,and then the convolution structure is used to extract the spatial information.Then,the feature maps of the two branches with different resolutions are gradually fused and upsampled.The algorithm is verified on the Cityscapes dataset,and the segmentation precision reaches 65.79%,which is 3.11%higher than BiSeNet.