Learning Semantic Graphics Using Convolutional Encoder–Decoder Network for Autonomous Weeding in Paddy

Abstract

Weeds in agricultural farms are aggressive growers which compete for nutrition and other resources with the crop and reduce production. The increasing use of chemicals to control them has inadvertent consequences to the human health and the environment. In this work, a novel neural network training method combining semantic graphics for data annotation and an advanced encoder–decoder network for (a) automatic crop line detection and (b) weed (wild millet) detection in paddy fields is proposed. The detected crop lines act as a guiding line for an autonomous weeding robot for inter-row weeding, whereas the detection of weeds enables autonomous intra-row weeding. The proposed data annotation method, semantic graphics, is intuitive, and the desired targets can be annotated easily with minimal labor. Also, the proposed “extended skip network” is an improved deep convolutional encoder–decoder neural network for efficient learning of semantic graphics. Quantitative evaluations of the proposed method demonstrated an increment of 6.29% and 6.14% in mean intersection over union (mIoU), over the baseline network on the task of paddy line detection and wild millet detection, respectively. The proposed method also leads to a 3.56% increment in mIoU and a significantly higher recall compared to a popular bounding box-based object detection approach on the task of wild–millet detection.

Deep Neural Network-Based System for Autonomous Navigation in Paddy Field

“…While [21], [22] used CNN-based semantic segmentation to discriminate crops, weeds and background, the actual lines of crop are not extracted. In our previous work we presented that CNN can directly be trained to learn the concept of a crop line using ''semantic graphics'' [23], as shown in Fig. 1.…”

A Multi-Scale Feature Fusion Method Based on U-Net for Retinal Vessel Segmentation

“…Simultaneously, the MSFFU-Net contained extended two skip connections: one was that each set of feature maps generated on the encoder path are concatenated to the corresponding feature maps on the decoder path; the other was that transferring of max pooling indices values from the encoder to the decoder to locate contour position information of multi-scale retinal vessel features for higher segmentation accuracy [ 32 ]. The feature maps of the upsampling operation were merged with the corresponding output feature maps of the two extended skip modules [ 33 ], as shown in Figure 6 .…”

DAM: Hierarchical Adaptive Feature Selection Using Convolution Encoder Decoder Network for Strawberry Segmentation

“…In some regions, labor cost makes up more than half of the total production cost, e.g., 60% in Norway ( Xiong et al, 2019 ). Furthermore, there is a decline in interest of joining the agriculture industry among the new generation of workers ( Adhikari et al, 2019 ). Under all these challenges the food industry must keep up with the demands of the ever-growing population.…”

“…Due to outstanding performances of DCNNs in computer vision tasks, robotics and unmanned systems are now faster and more reliable than ever. Which in turn has allowed their adoption into many real-life applications like the detection of crop rows, weeds, and seeding beds in fields of maize and rice ( Guerrero et al, 2017 ; Adhikari et al, 2019 ; Ma et al, 2019 ).…”