Clothing color analysis method and system, storage medium and computer equipment

Abstract

The invention provides a clothing color analysis method, which comprises the following steps of: determining a human body target in a to-be-detected image according to a target detection technology, and intercepting a human body region image of the human body target; performing image segmentation processing on the human body region image to obtain a segmented clothing region map; obtaining a color model of the clothing area map through color space modeling; density clustering is carried out in the color model through a Mean Shift algorithm, and the total number of pixels of each color category is counted based on a density clustering result; and according to a statistical result of the total number of the pixels, determining the color category with the highest numerical value in the total number of the pixels as a clothing color. The invention further provides a clothing color analysis system, a storage medium and computer equipment. Therefore, the clothing color identification accuracy can be effectively improved.

Description

technical field [0001] The invention relates to the technical field of detection and identification, in particular to a clothing color analysis method, system, storage medium and computer equipment. Background technique [0002] In the prior art, there are mainly the following schemes for clothing color recognition: image color mean method, color value pixel statistics method, neural network color classification method for images, and K-Means algorithm clustering method. [0003] However, in practical applications, the above methods have the following disadvantages: in the detection process, the image color cannot be averaged, so when the color is not a pure color, the result obtained by using the image color average method is uncertain. Although the color value pixel statistics method can obtain the color value with the most pixels, the range of color values ​​to be extracted is not easy to determine. Because the color is ever-changing, the color value on the image is any ...

AI Technical Summary

Problems solved by technology

[0003] However, in practical applications, the above methods have the following disadvantages: in the detection process, the image color cannot be averaged, so when the color is not a pure color, the result obtained by using the image color average method is uncertain

Although the color value pixel statistics method can obtain the color value with the most pixels, the range of color values ​​to be extracted is not easy to determine

Because the color is ever-changing, the color value on the image is any value between 0-255, which is a linear range; therefore, the detection accuracy of the neural network color classification method for images is not high in practical applications, and when in the same image When there are multiple colors, color classification cannot be performed accurately

The K-Means algorithm needs to determine the total number of categories for the clustering method, but it is impossible to determine which colors are in the image.

[0004] In summary, there are many problems in the actual use of existing methods, so it is necessary to improve

Images