An Image Segmentation Algorithm Based on a Local Region Conditional Random Field Model

Xiao Jiang,Shuaishuai Lv,Haibin Yu

doi:10.4236/ijcns.2020.139009

Xiao Jiang, Shuaishuai Lv + Show 1 more

Open Access

https://doi.org/10.4236/ijcns.2020.139009

Copy DOI

Abstract

To reduce the computation cost of a combined probabilistic graphical model and a deep neural network in semantic segmentation, the local region condition random field (LRCRF) model is investigated which selectively applies the condition random field (CRF) to the most active region in the image. The full convolutional network structure is optimized with the ResNet-18 structure and dilated convolution to expand the receptive field. The tracking networks are also improved based on SiameseFC by considering the frame relations in consecutive-frame traffic scene maps. Moreover, the segmentation results of the greyscale input data sets are more stable and effective than using the RGB images for deep neural network feature extraction. The experimental results show that the proposed method takes advantage of the image features directly and achieves good real-time performance and high segmentation accuracy.

Highlights

In the past twenty years, deep convolutional neural networks have gradually become a powerful tool for analysing images in various computer vision fields [1] [2] [3] [4]
By analysing the segmentation results in the convolution neural network, we found that the segmentation results of the fully connected conditional random field approach are not much improved in areas such as the sky, the road and buildings compared to the results obtained by only traditional convolutional neural networks
An improved conditional random field model application and a new region tracking application are proposed in this paper

Summary

Introduction

In the past twenty years, deep convolutional neural networks have gradually become a powerful tool for analysing images in various computer vision fields [1] [2] [3] [4]. Convolutional neural networks have achieved good results on image semantic segmentation tasks [2] [3]. Semantic image segmentation plays significant roles in the operation of autonomous vehicles and UAVs as well as wearable devices. Semantic segmentation for autonomous vehicle applications is different from general semantic segmentation because it must process video signals, which are acquired episodically by the car’s cameras. Semantic segmentation of images is concerned with the accuracy of the segmentation system but must achieve real-time performance in autonomous vehicle situations

Objectives

Results

Conclusion