The invention discloses a rectus abdominis separation measurement method based on key point detection, and relates to the technical field of ultrasonic
image processing, and the method comprises the steps: collecting an ultrasonic image; the ultrasonic image comprises a rectus abdominis area, real coordinates of
muscle tail ends, connected with an abdominal
white line, of the left side and the right side of rectus abdominis are marked, and a
data set is constructed; the ultrasonic image is preprocessed; creating a thermodynamic diagram according to the real coordinates; constructing a rectus abdominis
muscle end prediction model; based on the
data set, training the rectus abdominis
muscle end prediction model; predicting muscle end positions, connected with the abdominal
white line, of the left and right sides of the rectus abdominis in the
ultrasound image; based on the predicted muscle
tail end positions, connected with the abdominal
white line, of the left side and the right side of the rectus abdominis, the rectus abdominis separation distance is obtained. According to the method, a
probability estimation method is used, and meanwhile, an offset value is predicted for accurate positioning; and the Mobi leNet is used as the basic framework of the
backbone network, so that the equipment compatibility is better.