Measurement, Journal Year: 2024, Volume and Issue: 239, P. 115411 - 115411
Published: July 29, 2024
Language: Английский
Measurement, Journal Year: 2024, Volume and Issue: 239, P. 115411 - 115411
Published: July 29, 2024
Language: Английский
Measurement, Journal Year: 2025, Volume and Issue: unknown, P. 116918 - 116918
Published: Feb. 1, 2025
Language: Английский
Citations
1Agriculture, Journal Year: 2024, Volume and Issue: 14(10), P. 1846 - 1846
Published: Oct. 19, 2024
The wheat harvesting boundary line is vital reference information for the path tracking of an autonomously driving combine harvester. However, unfavorable factors, such as a complex light environment, tree shade, weeds, and stubble color interference in field, make it challenging to identify harvest accurately quickly. Therefore, this paper proposes recognition model based on MV3_DeepLabV3+ network framework, which can quickly complete identification environments. uses lightweight MobileNetV3_Large backbone LeakyReLU activation function avoid neural death problem. Depth-separable convolution introduced into Atrous Spatial Pyramid Pooling (ASPP) reduce complexity parameters. cubic B-spline curve-fitting method extracts line. A prototype harvester was built, field tests were conducted. test results show that proposed achieves segmentation accuracy 98.04% unharvested regions environments, with IoU 95.02%. When travels at 0~1.5 m/s, normal speed operation, average processing time pixel error single image are 0.15 s 7.3 pixels, respectively. This could achieve high fast speed. provides practical autonomous operation
Language: Английский
Citations
5Measurement, Journal Year: 2024, Volume and Issue: unknown, P. 115729 - 115729
Published: Sept. 1, 2024
Language: Английский
Citations
3Measurement, Journal Year: 2024, Volume and Issue: 237, P. 115072 - 115072
Published: June 12, 2024
A site-specific weed detection and classification system was implemented with a stereoscopic video camera to reduce the adverse effects of chemical herbicides in rice field. computer vision meta-heuristic hybrid NN-ICA classifier were used accurately discriminate between two varieties plants, under either natural light (NLC) or controlled conditions (CLC). Preprocessing, segmentation, matching procedures performed on images coming from right left channels. Most discriminant features selected average, arithmetic geometric, using NN-PSO algorithm. Accuracy results stereo NLC 85.71 % for mean (AM) 85.63 geometric (GM), test set. At same time, accuracy CLC reached 96.95 AM case 94.74 GM case, being consistently higher than those NLC.
Language: Английский
Citations
2Measurement, Journal Year: 2024, Volume and Issue: 239, P. 115411 - 115411
Published: July 29, 2024
Language: Английский
Citations
2