Image color correction method for directly correcting RGB value by adopting neural network

Abstract

The invention discloses an image color correction method for directly correcting RGB values by adopting a neural network, and belongs to the technical field of aerodynamic experimental measurement, and the method comprises the following steps: 1) establishing an image RGB color correction model based on a neural network algorithm; 2) taking the color calibration card as a reference color, and training parameters of a neural network color correction model; and 3) correcting the RGB value of the new image: taking the RGB value of the image shot by the camera as the input of the neural network, and obtaining the RGB value of the image after color correction at the output end of the neural network. According to the method disclosed by the invention, the problems of complex step sequence and information loss in a transformation process when an image color correction method is applied to an SSLC coating experimental measurement technology in the prior art are avoided. The method is simple and direct, is especially suitable for the SSLC coating measurement technology, and can also be used for other occasions requiring accurate color measurement.

Description

technical field [0001] The invention belongs to the technical field of aerodynamic experiment measurement, and relates to an image color calibration method, in particular to an image color calibration method using a neural network to directly correct RGB values. Background technique [0002] Wall friction is an important parameter in the field of aerodynamics. Accurate measurement of wall friction is of great significance in the fields of boundary layer theory research, flow control effect evaluation, and aircraft drag reduction research. Shear-sensitive liquid crystal (SSLC) coating technology is a non-contact method for measuring the wall friction vector field. The principle of this method is to spray a layer of SSLC coating (thickness is about 10um) on the surface to be tested. When the SSLC coating is subjected to the friction of the airflow, it will show different colors in different directions. The color displayed in the direction is analyzed and processed, and the fr...

AI Technical Summary

Problems solved by technology

Since the SSLC measurement technology needs to calculate the hue (Hue) value of the color according to the RGB components of the color for further analysis and processing and to solve the friction vector field, if the aforementioned traditional color calibration method is used, it is necessary to transform the color of the calibrated CIELAB space To the color of RGB space, but the color of CIELAB space cannot be directly converted to RGB color, it needs to be converted to CIEXYZ color first, and then converted to RGB color, which is more cumbersome

In addition, the CIELAB color space has a larger range than the RGB color space, and this transformation will inevitably lead to information loss

Images