The Use of a Stereovision System in Shape Detection of the Side Surface of the Body of the Mining Machine Working Unit

Abstract

Ensuring the compliance of the finished product with the project during the manufacturing of cutting heads/drums of the mining machines, largely determines the efficiency of rock mining, especially hard-to-cut rocks. The manufacturing process of these crucial elements of cutting machines is being robotized in order to ensure high accuracy and repeatability. This determines, among others the need to assess in real-time the degree of the approach of pick holders positioned by the industrial robot to the side surface of the working unit of the cutting machine in their target position. This problem is particularly important when in the manufacturing process are used the bodies of decommissioned cutting heads/drums, from which old pick holders have been removed. The shape and external dimensions of these hulls, unless they are subjected to regeneration, may differ quite significantly from the nominal ones. The publication, on the example of a road header cutting head, presents the procedure for automatically identifying and indexing markers displayed on its side surface, recorded on measuring photos by two digital cameras of a 3D vision system. Experimental research of the developed method was carried out using the KUKA VisionTech vision system installed on the test stand in the robotics laboratory of the Department of Mining Mechanization and Robotization at the Faculty of Mining, Safety Engineering and Industrial Automation of the Silesian University of Technology. Data processing was carried out in the Matlab environment using the libraries of the Image Processing Toolbox. The functions provided in this library were used in the developed algorithm, implemented in the software. This algorithm allows automatic identification of markers located in the images of the side surface of the cutting head. This is the basis for determining their location in space. The publication presents a method of segmenting images recorded by cameras into homogeneous areas. The method of separating interesting areas from the image by comparison to the pattern was presented. Also shown is the method of the automatic numbering of mutually matching pairs of markers on photos from two cameras included in the vision system depending on the spatial orientation of the marker grid in the measuring images.

Validation of a Method for Measuring the Position of Pick Holders on a Robotically Assisted Mining Machine’s Working Unit

“…The idea behind the measurement is to determine the spatial position of a grid of points projected onto the cutting head side surface with the use of a laser fitted with a diffraction grid (51 × 51 points). This projection device is mounted on the robotic arm that positions the pick holder [22,23]. A two-image method (stereophotogrammetry) is used for determining the spatial position.…”

“…Figure23. Distribution maps of the distances of measurement points from the cutting head side surface obtained by structural light scanning (reference method): (a) view of measurement points from the outer side of the cutting head side surface, (b) view of measurement points from the inner side of the cutting head side surface, and (c) zoomed-in view of the area of the largest deviation (inner side of cutting head side).…”

Applications of Machine Vision in Coal Mine Fully Mechanized Tunneling Faces: A Review

“…For instance, [95] discussed the possibility of utilizing a stereovision system to calculate the distance between the pick holder base and the roadheader cutting head side surface. Similarly, [48] [96] presented a method for detecting the side surface shape of the cutting head. They also determined the boom [97] and cutting head [98] vibrations of the roadheader during cutting based on an analysis of the time-lapse pictures of the recorded footage obtained from high-speed cameras using dedicated TEMA 3D software.…”