Fastening tool system and control method thereof
A control method of a fastening tool for fastening a component part includes photographing a fastening portion of the component part through a camera portion mounted on the fastening tool. The control method includes pre-processing that rotates an input image, which has been photographed by the camera portion, to match with representative model image. The control method includes estimating the fastening portion through an inference process through a convolutional neural network (CNN) algorithm-based image classification work on a video input image of a same fastening portion finished with the pre-processing work. The control method also includes setting a torque value that is matched with a recognized fastening portion when a probability value of the fastening portion in the inference process exceeds a predetermined reference ratio.
1. A control method of a fastening tool for fastening a component part, the control method comprising:
photographing a fastening portion of the component part through a camera portion mounted on the fastening tool;
pre-processing that rotates an input image photographed by the camera portion to match with representative model image;
estimating the fastening portion through an inference process through a convolutional neural network (CNN) algorithm-based image classification work on a video input image of a same fastening portion finished with the pre-processing work; and
setting a torque value matched with a recognized fastening portion when a probability value of the fastening portion in the inference process exceeds a predetermined reference ratio,
wherein the pre-processing includes
masking an outer portion of a circle based on a center of the input image,
calculating a slope angle according to a statistical value of slope distribution for each region of the input image obtained through histogram of gradient (HOG) algorithm, and
performing matching by rotating the input image based on the slope angle.
2. The control method of claim 1 , wherein the pre-processing further comprises:
adjusting brightness of the input image through histogram equalization work; and
masking a tool socket image existing in the input image.
3. A control method of a fastening tool for fastening a component part, the control method comprising:
photographing a fastening portion of the component part through a camera portion mounted on the fastening tool;
pre-processing that rotates an input image photographed by the camera portion to match with representative model image;
estimating the fastening portion through an inference process through a convolutional neural network (CNN) algorithm-based image classification work on a video input image of a same fastening portion finished with the pre-processing work; and
setting a torque value matched with a recognized fastening portion when a probability value of the fastening portion in the inference process exceeds a predetermined reference ratio,
wherein estimating the fastening portion includes
accumulating and averaging probability information output from a previous frame of the video input image, and
selecting a fastening portion of a class having a highest probability value among class information of the fastening portion and probability values consecutively output for the video input image during the inference process.
4. The control method of claim 3 , wherein, in the setting of the torque value, when the highest probability value is below the reference ratio, the torque value is manually set, and images photographed for a predetermined period of time immediately prior to currently performed fastening work are stored as additional learning data.