Target search method and robot for visual robot

Abstract

The invention provides a target searching method and a robot of a visual robot, and relates to the technical field of robot searching. The method includes the steps of: acquiring search target information; the search target information includes target features and the target type; starting a visual camera to capture a field of view image, according to the size requirements of the detection image size of the target detection neural network model corresponding to the target type. , preprocess the visual field image to generate a detection image, and input the detection image into the aforementioned corresponding target detection neural network model for target detection; the detection image is stitched from multiple images with different resolutions and related to the visual field image obtain the target detection result, and determine whether there is a search target in the target detection result according to the aforementioned target characteristics. On the basis of ensuring the real-time accuracy of far and near target searching, the present invention reduces the requirement on equipment computing power for far and near target searching.

Description

technical field [0001] The invention relates to the technical field of robot search. Background technique [0002] The search for targets by intelligent mobile robots is widely used in search and rescue services, scene monitoring and object finding, target detection and other fields, and has broad application prospects. [0003] The search for targets by mobile robots may involve multiple fields such as computer vision and pattern recognition. Among them, for computer vision, the purpose is to simulate the human visual mechanism, calculate the importance of information in the visual scene, and extract the salient features or target object features of interest in the image. The process of robot target search based on vision usually includes image acquisition, target detection and target confirmation. With the rapid development of artificial intelligence and deep learning technology, object detection methods based on convolutional neural network (CNN) algorithms have been wi...

AI Technical Summary

Problems solved by technology

[0004] However, on the one hand, although the convolutional neural network algorithm improves the real-time accuracy of tracking, because the convolutional neural network-based target detection algorithm usually contains a large number of calculation-intensive operations, the requirements for real-time detection computing power and bandwidth are very high.

Images