Shoe sole size classification method based on machine vision

Abstract

The invention discloses a shoe sole size classification method based on machine vision, which comprises the following steps: (1) collecting standard shoe sole images of different sizes, extracting shoe sole image contours, calculating standard shoe sole parameters of different sizes, including a standard arch width, a standard palm width and a standard heel width, and storing the standard shoe sole parameters as standard parameters; (2) collecting to-be-classified sole images, extracting to-be-classified sole image contours, calculating to-be-classified sole parameters including to-be-classified arch widths, to-be-classified palm widths and to-be-classified heel widths, and assigning values to the to-be-classified parameters; (3) sequentially comparing the to-be-classified parameters withall stored standard parameters, and respectively calculating classification matching values of the to-be-classified parameters and the stored standard parameters; and (4) sorting the classification matching values, wherein the standard shoe sole size corresponding to the minimum classification matching value is the shoe sole size to be classified. Aiming at the problems of low accuracy and low efficiency of current machine vision sole size classification, the method provided by the invention can realize accurate and rapid classification of the sole size, is high in efficiency and saves the labor cost.

Description

technical field [0001] The invention relates to the field of intelligent detection of automatic production line products, in particular to a method for classifying shoe sole sizes based on machine vision. Background technique [0002] In the automatic production of the shoemaking industry, it is necessary to bond the sole and the upper to make finished shoes. Before the bonding process, the size of the sole should be detected and classified to prevent the mismatch between the size of the sole and the upper, resulting in defective products and resulting in production waste. . Traditional detection and classification methods include manual weighing classification and naked eye recognition classification. Manual weighing classification is to weigh the quality of the sole, and judge the size according to the quality of the sole. However, when the difference between the soles is half a size, the difference in quality is small, which often leads to misjudgment. Naked-eye recogni...

AI Technical Summary

Problems solved by technology

Manual weighing classification is to weigh the quality of the sole and judge the size according to the quality of the sole. However, when the difference between the soles is half a size, the quality difference is small, which often leads to misjudgment.

Naked-eye recognition classifies soles of different sizes by reading the size numbers on the surface of the soles. However, due to the shading on the soles, the size numbers are not clear, and the human eye is prone to fatigue after long-term work, which often leads to wrong size judgments.

To sum up, the traditional method has the following disadvantages: 1. Slow speed and low efficiency; 2. The recognition accuracy is not high; 3. High labor cost

Machine vision has been used in many inspection fields, and there are also application precedents in size inspection, such as the machine vision inspection method for recognizing the digital code of the sole, and the machine vision inspection method for judging the size of the sole by the size of the sole, but the sole Curls will occur, making it difficult to identify the digital code on the sole, and the area of ​​the sole will also change greatly. It is difficult to meet the requirements of accurate and fast classification of the sole size. Therefore, how to achieve accurate and fast size classification of the sole has troubled technicians for many years. question

Images