Face Sketch Generation Using Evolutionary Computing

Abstract

In this paper, an evolutionary genetic algorithm is used to generate face sketch from the face description. Face sketch generation without face image is extremely important for the law enforcement agencies. The genetic algorithm is used for generating face sketch through several iterations of the algorithm. The face image description is captured through graphical user interface just by clicking options for each face features. Face features are used to extract face images and generate initial population for the genetic algorithm. Genetic operators such as selection, crossover and mutation are used for next generation of the population. The Genetic algorithm cycle is repeated until the user is satisfied with face sketch generated. The novelty of the paper includes face sketch generation from face image description. The result shows that evolutionary based technique for sketch generation produces the desired face sketch.

A New Framework for Matching Forensic Composite Sketches With Digital Images

“…A robust framework called local feature based discriminant analysis was proposed by Klare et al, (2010) .The proposed framework outperforms the other existing approaches in case of matching viewed sketches. An evolutionary genetic algorithm was proposed by Bansode et al, (2016) which generates face sketch from face description.…”

Face Photograph Recognition via Generation from Sketches using Convolutional Neural Networks

“…Tang and Wang [11] have compared face sketch images with real images by suggesting an automatic image retrieval system in the CUHK data set. Bansode and Sinha [12] proposed a genetic algorithm-based image generation system from color image to face sketch image.…”