A High-Density Crowd Counting Method Based on Convolutional Feature Fusion
| Autor: | Weiqun Wu, Hongling Luo, Hong Xiang, Jun Sang, Zhongyuan Wu, Qian Zhang, Zhili Xiang |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2018 |
| Předmět: |
Computer science
02 engineering and technology lcsh:Technology Convolutional neural network Image (mathematics) lcsh:Chemistry Crowds density map 0202 electrical engineering electronic engineering information engineering feature fusion crowd counting General Materials Science lcsh:QH301-705.5 Instrumentation Crowd counting Fluid Flow and Transfer Processes Fusion lcsh:T business.industry Process Chemistry and Technology General Engineering 020207 software engineering Pattern recognition Sample (graphics) Convolutional Neural Network (CNN) lcsh:QC1-999 Computer Science Applications lcsh:Biology (General) lcsh:QD1-999 lcsh:TA1-2040 020201 artificial intelligence & image processing Artificial intelligence Deconvolution lcsh:Engineering (General). Civil engineering (General) Joint (audio engineering) business lcsh:Physics |
| Zdroj: | Applied Sciences Volume 8 Issue 12 Applied Sciences, Vol 8, Iss 12, p 2367 (2018) |
| ISSN: | 2076-3417 |
| DOI: | 10.3390/app8122367 |
| Popis: | In recent years, the trampling events due to overcrowding have occurred frequently, which leads to the demand for crowd counting under a high-density environment. At present, there are few studies on monitoring crowds in a large-scale crowded environment, while there exists technology drawbacks and a lack of mature systems. Aiming to solve the crowd counting problem with high-density under complex environments, a feature fusion-based deep convolutional neural network method FF-CNN (Feature Fusion of Convolutional Neural Network) was proposed in this paper. The proposed FF-CNN mapped the crowd image to its crowd density map, and then obtained the head count by integration. The geometry adaptive kernels were adopted to generate high-quality density maps which were used as ground truths for network training. The deconvolution technique was used to achieve the fusion of high-level and low-level features to get richer features, and two loss functions, i.e., density map loss and absolute count loss, were used for joint optimization. In order to increase the sample diversity, the original images were cropped with a random cropping method for each iteration. The experimental results of FF-CNN on the ShanghaiTech public dataset showed that the fusion of low-level and high-level features can extract richer features to improve the precision of density map estimation, and further improve the accuracy of crowd counting. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|