An automated method for stem diameter measurement based on laser module and deep learning.

BACKGROUND: Measuring stem diameter (SD) is a crucial foundation for forest resource management, but current methods require expert personnel and are time-consuming and costly. In this study, we proposed a novel device and method for automatic SD measurement using an image sensor and a laser module. Firstly, the laser module generated a spot on the tree stem that could be used as reference information for measuring SD. Secondly, an end-to-end model was performed to identify the trunk contour in the panchromatic image from the image sensor. Finally, SD was calculated from the linear relationship between the trunk contour and the spot diameter in pixels. RESULTS: We conducted SD measurements in three natural scenarios with different land cover types: transitional woodland/shrub, mixed forest, and green urban area. The SD values varied from 2.00 cm to 89.00 cm across these scenarios. Compared with the field tape measurements, the SD data measured by our method showed high consistency in different natural scenarios. The absolute mean error was 0.36 cm and the root mean square error was 0.45 cm. Our integrated device is low cost, portable, and without the assistance of a tripod. Compared to most studies, our method demonstrated better versatility and exhibited higher performance. CONCLUSION: Our method achieved the automatic, efficient and accurate measurement of SD in natural scenarios. In the future, the device will be further explored to be integrated into autonomous mobile robots for more scenarios.

Deep Learning with Taxonomic Loss for Plant Identification.

Plant identification is a fine-grained classification task which aims to identify the family, genus, and species according to plant appearance features. Inspired by the hierarchical structure of taxonomic tree, the taxonomic loss was proposed, which could encode the hierarchical relationships among multilevel labels into the deep learning objective function by simple group and sum operation. By training various neural networks on PlantCLEF 2015 and PlantCLEF 2017 datasets, the experimental results demonstrated that the proposed loss function was easy to implement and outperformed the most commonly adopted cross-entropy loss. Eight neural networks were trained, respectively, by two different loss functions on PlantCLEF 2015 dataset, and the models trained by taxonomic loss led to significant performance improvements. On PlantCLEF 2017 dataset with 10,000 species, the SENet-154 model trained by taxonomic loss achieved the accuracies of 84.07%, 79.97%, and 73.61% at family, genus and species levels, which improved those of model trained by cross-entropy loss by 2.23%, 1.34%, and 1.08%, respectively. The taxonomic loss could further facilitate the fine-grained classification task with hierarchical labels.

Visual Cortex Inspired CNN Model for Feature Construction in Text Analysis.

Recently, biologically inspired models are gradually proposed to solve the problem in text analysis. Convolutional neural networks (CNN) are hierarchical artificial neural networks, which include a various of multilayer perceptrons. According to biological research, CNN can be improved by bringing in the attention modulation and memory processing of primate visual cortex. In this paper, we employ the above properties of primate visual cortex to improve CNN and propose a biological-mechanism-driven-feature-construction based answer recommendation method (BMFC-ARM), which is used to recommend the best answer for the corresponding given questions in community question answering. BMFC-ARM is an improved CNN with four channels respectively representing questions, answers, asker information and answerer information, and mainly contains two stages: biological mechanism driven feature construction (BMFC) and answer ranking. BMFC imitates the attention modulation property by introducing the asker information and answerer information of given questions and the similarity between them, and imitates the memory processing property through bringing in the user reputation information for answerers. Then the feature vector for answer ranking is constructed by fusing the asker-answerer similarities, answerer's reputation and the corresponding vectors of question, answer, asker, and answerer. Finally, the Softmax is used at the stage of answer ranking to get best answers by the feature vector. The experimental results of answer recommendation on the Stackexchange dataset show that BMFC-ARM exhibits better performance.