Method for discriminating vehicle colors based on convolution neural network model

Abstract

The embodiment of the invention discloses a method for discriminating vehicle colors based on a convolution neural network model, and the method comprises the steps: obtaining an image of a vehicle; recognizing the license plate information in the vehicle image, wherein the license plate information comprises the length, width and position of a license plate; determining a vehicle head region and a vehicle head image of a vehicle head region according to the information of the license plate; converting the vehicle head image into an image at a YUV format, and obtaining a vehicle head YUV image; extracting Y-component data, U-component data and V-component data from the vehicle head YUV image, carrying out the recombination of the Y-component data, U-component data and V-component data, and obtaining a vehicle head YUYV image in a preset size; training the convolution neural network model; inputting the vehicle head YUYV image into the convolution neural network model; and determining the color of the head of a vehicle according to an output result of the convolution neural network model. The embodiment of the invention provides a method for recognizing vehicle colors based on deep learning.

Description

technical field [0001] The invention relates to the field of image processing, in particular to a method and device for discriminating vehicle colors based on a convolutional neural network model. Background technique [0002] In the prior art, at present, the color recognition of vehicles is mainly carried out by comparing the distribution of color components counted by statistical methods. The advantage of this method is that the recognition speed is fast, but the disadvantage is that the recognition rate is not high and it is easily affected by the environment. Used for pre-identification or coarse classification. Another method is to use the traditional pattern learning method to classify the color of the car through the color feature model of the training sample. The training efficiency of this method is not high. Contents of the invention [0003] Embodiments of the present application provide a method and device for identifying vehicle color based on a convolutiona...

AI Technical Summary

Problems solved by technology

[0002] In the prior art, at present, the color recognition of vehicles is mainly carried out by comparing the distribution of color components counted by statistical methods. The advantage of this method is that the recognition speed is fast, but the disadvantage is that the recognition rate is not high and it is easily affected by the environment. For pre-identification or coarse classification

Another method is to use the traditional pattern learning method to classify the color of the car through the color feature model of the training sample. The training efficiency of this method is not high.

Images