Laser & Optoelectronics Progress, Volume. 56, Issue 9, 091003(2019)
Low-Parameter Real-Time Image Segmentation Algorithm Based on Convolutional Neural Network
Fig. 1. Convolution kernel. (a) Classical convolution kernel; (b) dilated convolution kernel Rrate=2; (c) dilated convolution kernel Rrate=3
Fig. 2. Atrous-Fire modular structure
Fig. 3. Dilated convolution kernel and initial characteristic graphs. (a) Sawtooth structure convolution kernel; (b) no grid feature graph; (c) grid feature graph
Fig. 4. Network structure of Atrous-squeezeseg
Fig. 5. Training loss value curves
Fig. 6. Validation loss value curves
Fig. 7. Effect comparison of ADE20K. (a) Original images; (b) ground truth; (c) proposed algorithm; (d) Squeezeseg+FCN; (e) VGG16+FCN; (f) SqueezeNet+FCN; (g) without dilated; (h) without BN
Fig. 8. Effect comparison of PASCAL VOC. (a) Original images; (b) ground truth; (c) proposed algorithm; (d) Squeezeseg+FCN; (e) VGG16+FCN; (f) SqueezeNet+FCN; (g) without dilated; (h) without BN
|
|
|
Get Citation
Copy Citation Text
Guanghong Tan, Jin Hou, Yanpeng Han, Shuo Luo. Low-Parameter Real-Time Image Segmentation Algorithm Based on Convolutional Neural Network[J]. Laser & Optoelectronics Progress, 2019, 56(9): 091003
Category: Image Processing
Received: Oct. 22, 2018
Accepted: Nov. 30, 2018
Published Online: Jul. 5, 2019
The Author Email: Hou Jin (jhou@swjtu.edu.cn)