GENERAL INFORMATION 
...........................

1. Dataset title:
Proposal and Implementation of a Procedure for Compliance Recognition of Objects with Smart Tactile Sensors - Dataset

2. Authors:
Raúl Lora-Rivera
Óscar Oballe-Peinado
Fernando Vidal-Verdú

3. Author contact information: 
Raúl Lora-Rivera - raul.lora@uma.es


METHODOLOGICAL INFORMATION 
.................................

1. Description of the methods for collection/generation of data: 
The dataset was collected using a custom-designed smart tactile sensor suite integrated into a Cartesian robotic platform, which performed controlled compression and decompression cycles on 42 objects varying in shape and compliance (see the paper document for more information). 

Lora-Rivera, R.; Oballe-Peinado, Ó.; Vidal-Verdú, F. Proposal and Implementation of a Procedure for Compliance Recognition of Objects with Smart Tactile Sensors. Sensors 2023, 23, 4120. https://doi.org/10.3390/s23084120

2. Software or instruments needed to interpret the data:
The data is supplied in .mat format, making it accessible through MATLAB software.


FILE OVERVIEW 
----------------------

This document includes two attached .mat files containing the sequence of image moments derived from tactile sensor data on the finger and palm of a robotic hand (for further details, refer to the associated paper). 


DATA-SPECIFIC INFORMATION:
-------------------------------------------

1. Name File: fingerImageMoment.mat"

1.1. Variables list:
This is a .mat file containing the first six sequence image moments for the finger and the 42 object classes defined in the paper. 
Specifically, in the top level of the .mat structure, there is a 42x1 size cell array, where each entry corresponds to an object class.
Inside each of these entries, there is another level with a cell array structure of N x 1 size. N is the number of iterations carried out for the corresponding object class. 
Then, inside each iteration, there is another cell array of M x C size. M is the length of the image sequence moment of acquired data, and C goes from 1 to 6, where:

	--> C = 1, is the M_{0,0} image moment sequence.
	--> C = 2, is the M_{0,1} image moment sequence.
	--> C = 3, is the M_{1,0} image moment sequence.
	--> C = 4, is the M_{0,2} image moment sequence.
	--> C = 5, is the M_{2,0} image moment sequence.
	--> C = 6, is the M_{1,1} image moment sequence.		


1. Name File: palmImageMoment.mat

1.1. Variables list:
This is a .mat file containing the first six sequence image moments for the palm and the 42 object classes defined in the paper. 
Specifically, in the top level of the .mat structure, there is a 42x1 size cell array, where each entry corresponds to an object class.
Inside each of these entries, there is another level with a cell array structure of N x 1 size. N is the number of iterations carried out for the corresponding object class. 
Then, inside each iteration, there is another cell array of M x C size. M is the length of the image sequence moment of acquired data, and C goes from 1 to 6, where:

	--> C = 1, is the M_{0,0} image moment sequence.
	--> C = 2, is the M_{0,1} image moment sequence.
	--> C = 3, is the M_{1,0} image moment sequence.
	--> C = 4, is the M_{0,2} image moment sequence.
	--> C = 5, is the M_{2,0} image moment sequence.
	--> C = 6, is the M_{1,1} image moment sequence.