An eye- tracking technology and MLP-based color matching design method

Abstract

Images are a significant source of inspiration for designers to carry out the color design. However, the absence of animated images in the product color design can create confusion for designers. To translate the colours of the animated images into product colours, this work used eye-tracking technology to aid colour extraction and the multilayer perceptron neural network (MLP) algorithm to train a product colour decision model to filter the best product colour schemes. Firstly, eye tracking technology is used to collect the distribution of hotspots of the subject while viewing the animated images. Based on the distribution of eye-tracking hotspots, the most interesting animated colours were extracted. Then, the MLP is applied to train a colour decision model for children's shopping cart products, and the colour decision model is used to filter the optimal solution for the product colour, and finally the colour design is completed from the animated colour to the three-colour children's shopping cart product. Experimental results show that the color extraction based on the eye-tracking technology and the color scheme screening based on the intelligent algorithm can realize the effective conversion from animated image colors to product colors. This work proposes a color scheme design method from animations to products, which further expands the image color sources in product color design and can accurately find the color scheme that matches the animated image and the product.