ABSTRACT The tibial‐calcaneal angle (TCA) is a crucial biomechanical parameter used to assess foot alignment and classify rearfoot positions such as varus, rectus and valgus. Advances in plantar pressure sensing and deep learning (DL) have improved the accuracy of foot posture assessment, yet current methods rarely integrate plantar pressure with TCA, limiting precision. This manuscript presents an automated system for rearfoot classification based on plantar pressure mapping. The proposed method combines a 32 32 force‐sensitive resistor (FSR) sensor array with a convolutional neural network (CNN) to improve classification accuracy, reducing dependence on manual and radiographic techniques. Plantar pressure data were collected from 100 participants, and the CNN was trained to classify four alignment categories: rectus–rectus, rectus–valgus, valgus–valgus, and valgus–rectus. In this notation, the first term denotes the left foot and the second the right. Rectus indicates neutral alignment, while valgus refers to outward heel deviation. The model achieved 90.00% accuracy, showing strong generalization across categories. The valgus–rectus class exhibited the highest recall, approximately 93.75% with 97.58% specificity, while valgus–valgus reached 88.57% recall. Overall, the proposed system demonstrates high accuracy and robust performance, representing a clinically viable alternative to conventional image‐based methods.

To achieve successful autonomous navigation, a vision system that provides continuous and precise three-dimensional data of the system’s surroundings is always needed. An indispensable step in the acquisition of reliable data is the calibration of the system, preferably with a time-efficient and low-complexity approach. In this paper, a robust and efficient calibration method is proposed for the information fusion of a stereo vision system and a Technical Vision System. The proposed methodology achieves an error of xe= 6.9401 mm, ye= 8.0997 mm and ze= 15.1822 mm in 3.0202 seconds of processing time, as proven through experimental results.

non-contact 3D spatial coordinate measurement systems, based on optical laser scanning as technical vision systems (TVS) for signal processing, are essential methodologies for obtaining topographies in high-risk environments where human exploration is limited. However, these systems have limitations in resolution, particularly when addressing features such as surface texture and small curvatures at edges. Therefore, in this work, we propose the implementation of speckle pattern projection as a complementary innovative solution. Supported by the digital image correlation (DIC) methodology and the use of multivariate methods such as principal component analysis (PCA), we obtain results from different wall surfaces in a pipe prototype. Additionally, we analyze the behavior of the signal received by the 3D scanner sensor, which provides complementary information about the study surface. This demonstrates that the combination of speckle pattern projection and three-dimensional laser scanning is an additional tool for advanced detection of substance material during pipeline monitoring.

Stereo vision systems are well know depth estimation methods with a large number of applications such as automatic inspection, autonomous navigation, process control, etc. The functioning principle of these systems is the triangulation between the real-world surface point and its respective projections on the image planes of each camera. One of the key points in order to obtain accurate measurements on stereo vision systems are the calibration of extrinsic and intrinsic parameters. This is why the work of this paper focuses on a camera calibration method to correct the error generated by the lens distortion. The proposed method divides the image in quadrants and generates an equation for each quadrant to correct the error generated by the lens distortion. The performed experiment demonstrated an accuracy improvement using the calibration method compared to the measures taken without a calibration method.

This paper proposes an Artificial Neural Network (ANN) to accurately predict the real angles obtained by a Triangulation Vision System. The performance of the ANN is compared with the K-Nearest Neighbors algorithm from previous publications. For the experimentation it was necessary to create a database to train and prove both methods in different coordinates on a determinate area through the dynamic triangulation method. Afterwards, the root mean square error is calculated to obtain the accuracy of each algorithm. Finally, several laser scanning measurements were taken at different distances to analyze the measurement dispersion of both algorithms.

Laser vision systems have demonstrated to be useful in applications for autonomous navigation, structural health monitoring, manufacturing, reverse engineering, among others. A variant of these systems are the dynamic triangulation systems, where, these vision systems consist in a positioning laser, a scanning aperture and a fixed distance between them. The positioning laser points the laser beam over the surface to scan and it is detected by the scanning aperture. The purpose of this paper is to present the principle of operation of this system, the disadvantages when taking measures at different distances, and the implementation of the k-nearest neighbor algorithm (kNN) to solve these disadvantages.