Rapid object recognition method for mobile robot based on deep learning

Abstract

The invention discloses a rapid object recognition method for a mobile robot based on deep learning, and the method comprises the following steps: 1), the obtaining of a mobile image; 2), the preprocessing of image data; 3), the extracting of the features of the image; 4), image prediction output; 5), environment constrained optimization; 6), image recognition output. Through the unified integration of detection and recognition results, the method overcomes the shortcomings of the complexity and instability of a conventional object recognition system which needs to employ an object. Through a multilayer residual error network design and the generation of environment gravity constraint conditions, the method irons out the defect that an integrated object recognition system is poor in accuracy. The integration of detection and recognition tasks can guarantee the processing efficiency of a system, and improves the sensing capability of the robot in a moving process.

Description

technical field [0001] The invention relates to a fast object recognition method for a mobile robot, in particular to a fast object recognition method for a mobile robot based on deep learning. Background technique [0002] At present, there are object detection and recognition systems on the market. Generally, a set of object candidate areas is obtained through sliding window or object area recommendation technology, and then heuristic feature selection is used to identify the candidate areas in this set. The candidate region specifies the closest object class. [0003] According to the different features used by the classifier, the current detection system can be roughly divided into two types: the method based on deep learning; the method based on heuristic features. The scheme based on heuristic features generally represents the candidate area through empirical artificial design features, and based on the deep learning method, the object features are extracted through a...

AI Technical Summary

Problems solved by technology

[0006] In view of the decrease in operating efficiency caused by the separation of detection and recognition components in the object detection and recognition system, the functions of object detection prediction and recognition of its category are completed simultaneously. In order not to lose the accuracy of detection, a more expressive residual type multiple Layer-depth network improves the accuracy of integrated solutions and solves the shortcomings of existing technologies

Images