A material classification method based on joint sparse coding of tactile information of dexterous hands

Abstract

The invention relates to a material classification method based on joint sparse coding of dexterous hand tactile information, belonging to the technical field of material classification. The method includes: 1) collecting tactile information of an object used as a training sample; 2) according to the different materials of the training sample, Divide the training samples into i categories, grab each training sample, collect tactile information to obtain a tactile time series, and establish a training sample data set; 3) According to the obtained training sample data set, extract the characteristics of the training samples, and establish a tactile sequence dictionary Φ(D); 4) Grab the test sample object that needs to be classified to obtain the tactile time series of the test sample, and classify the material of each obtained tactile time series of the test sample to obtain the category of the test sample; 5 ) traverse the above step 4) for all test samples, and obtain the category of the material of each test sample. The invention realizes the material classification based on the tactile information on the basis of the joint sparse coding method, and improves the robustness and accuracy of the classification.

Description

technical field [0001] The invention belongs to the technical field of material classification, in particular to a material classification method based on joint sparse coding of dexterous hand tactile information. Background technique [0002] The BarrettHand dexterous hand of the BH8 series is a programmable multi-finger gripper, which has very high flexibility and can grasp target objects of different shapes, sizes and postures. The dexterous hand is composed of three fingers F1, F2, F3 and four tactile sensors of a palm S. Each sensor contains 24 tactile arrays, which can grasp objects and collect tactile information of the grasped objects to obtain tactile time series. [0003] In the existing object type recognition technology, there are many different methods of object type recognition technology, such as: the object type recognition technology based on heuristic rules, which uses simple information such as the shape, size, and proportion of the object to extract The ...

AI Technical Summary

Problems solved by technology

However, this method only extracts different frame images of objects as classification features, and does not apply similar methods to tactile information.

Images