GauTure Appears at Top International Academic Conference ICRA 2024

 ICRA (IEEE International Conference on Robotics and Automation) 2024, the top international academic conference on robotics, was held in Yokohama, Japan, attracting global researchers in the field of robotics, companies and scholars from all over the world to discuss the latest scientific advances and industrial achievements in the field of robotics. As the top conference in the field of robotics and automation, ICRA paper topics cover areas such as robotics, human-robot interaction, unmanned aerial systems, artificial intelligence, agricultural automation, behavioral trees, and big data analytics, among others.

Previously, GauTure's research result "AV4GAInsp: An Efficient Dual-Camera System for Identifying Defective Kernels of Cereal Grains" was accepted and published by IEEE Robotics and Automation Letters, a journal of the Institute of Electrical and Electronics Engineers. Automation Letters of the Institute of Electrical and Electronics Engineers (IEEE Robotics and Automation Letters), and was invited to participate in the conference. Dr. Lei Fan, one of the authors of the paper, was appointed by GauTure to attend the conference. Dr. Fan Lei reported the paper jointly participated by GauTure's R&D team members in the conference, and elaborated the multigrain collection device and multigrain algorithm (AV4GAInsp) made by GauTure's R&D team, during the period, Fan Lei publicized the achievement on the spot and exchanged ideas with the scholars, which was very fruitful. Congratulations to GauTure's paper accepted by IEEE Robotics and Automation Letters journal!

A machine vision-based automated grain appearance quality inspection system (AV4GAInsp)

    In this paper, we present a machine vision-based automated grain appearance quality inspection system (AV4GAInsp) designed to improve the efficiency and accuracy of grain quality inspection.The AV4GAInsp system consists of a multi-grain acquisition device P600 and a deep learning model-based grain analysis framework. Among them, the multi-grain acquisition device P600 unit is used to efficiently and automatically acquire high-quality appearance images of multiple types of multi-grains. The deep learning based analysis framework processes the captured images using a multi-stage artificial intelligence algorithm to analyze the imperfect grain types of the grains.

In addition, the paper constructs and publishes a large-scale multi-grain dataset, called GrainDet, which includes more than 140,000 images of three grains (wheat, sorghum, and rice). The effectiveness and performance of the AV4GAInsp system was verified through comprehensive experiments, with an average F1 score of 98.4% and an inspection efficiency more than 20 times higher than manual inspection. Multiple devices and previous conformance experiments by QC experts were also tested, and multiple sets of Kappa statistical tests confirmed the conformance between the system and human experts.

    It is expected that AV4GAInsp will effectively reduce the work pressure of inspectors and at the same time stimulate the enthusiasm for more in-depth research in the field of intelligent agriculture.

     By participating in this top conference, GauTure has demonstrated its strong strength in the field of scientific research. We promise that GauTure will continue to innovate and boldly expand the new realm of intelligent machine technology in the intelligent manufacturing application scenarios in the grain industry. We will actively play a leading role in the grain industry, stimulate unlimited creativity, and promote the development of the application of new scenarios of intelligent grain. GauTure is committed to providing all-round support for all links in the grain industry chain, including scene implementation, production cooperation and innovation incubation, in order to promote the intelligent upgrade and value growth of the whole industry.