Meal Recognition Based On Image Processing And The Feeding Plan Of Meal-aiding Robot

In recent years,with the continuous advancement of the intelligent information age,the level of technological development and artificial intelligence technology have continued to make major breakthroughs,and robotics has gradually entered all walks of life.In order to better solve the existing nursing problems of the elderly and the disabled,meal-aiding robots have gradually become a hot spot in the field of service robots.How to make catering robots better serve the elderly and the disabled has become the main direction of related technology development.Therefore,in this paper,in view of the existing problems of the existing food-assisting robots in the recognition and positioning of dishes,the use of computer vision technology to assist the robot in completing related tasks has a good guiding significance for the development of robotics.Based on computer vision technology,this paper proposes an intelligent system for classifying,identifying and locating food on a plate by combining image processing,object detection algorithms and binocular vision technology.The system mainly uses the Retinex-based image enhancement optimization algorithm and the variational model-based image filtering improvement algorithm to complete the preprocessing task of the image set,which greatly enhances the quality of the image and reduces the overall time consumption of preprocessing;From the AQ module to the feature extraction network and the MQ prediction layer,an improved YOLOv3 network model is designed.The model is used to train and predict the image set to achieve the task of classifying and identifying the food in the plate,which not only improves the accuracy of the target detection algorithm It also greatly speeds up the detection speed;uses binocular vision technology to establish a set of binocular ranging positioning system,using this system to complete the three-dimensional positioning task of the image in the dinner plate;according to the three-dimensional positioning of the three-dimensional model and the classification results of the target detection,Estimate the posture angle of the food,and provide complete information for the follow-up manipulator to grab the food.Completed the task of classification and recognition of the food in the plate,three-dimensional positioning and posture angle estimation tasks in the meal-aiding robot system.