An online tool wear detection system in dry milling based on machine vision

Tool wear is accelerated with the friction in the tool–workpiece contact during dry cutting. Tool changing early or late will affect the quality of tool and workpiece. An online and machine system vision-based is built to monitor tool condition in real time. MATLAB is used to compile the self-matching algorithm, which considers the features of interested targets on the flank face. Furthermore, a corresponding GUI is designed and encapsulated for both the bottom and flank edges. It is shown that the absolute value of the error on the maximum wear width is not more than 0.007 mm for the bottom edge. For the flank edge, the absolute value of the error is not more than 0.030 mm owing to the local highlighting interference. It is proved that the system can guarantee the quality of tool and workpiece and avoid unnecessary waste significantly. This platform can enhance the utilization of the tool in dry cutting.