Face recognition algorithm is identifying people’s name based on their face images. This subject is instrumental in people’s identification since people are easily photographed without any physical contact in comparison to other biometric methods. Consequently, people experience slight discomfort in presence of surveillance cameras. However, other biometric methods disturb people because they need to contact them physically. One of the essential extracting methods in face recognition is LBP (Local Binary Pattern). LBP was first proposed to describe image texture [1]. Then, it has been widely employed as a feature extracting method in many applications of image and video processing [2, 3, 4, 5, 6]. Regarding enhancement of LBP, various LBP methods were proposed. Most of these LBP extensions have a rigid structure to extract information from images. Meanwhile, image data can be exploited to enhance LBP functions. In this paper, LBP is modified to become a better feature extractor for various applications, in particular face recognition through exploitation of application data. Hence, the LBP coefficients are converted to chromosomes, possible solutions in the genetic algorithm. Meanwhile, the genetic operators such as cross-over and mutation operators make new chromosomes invalid. This paper proposes a constructive method to correct these defective chromosomes, particularly it rectifies chromosomes violating the summation constraint, vital to draw histograms. The paper is organized as follows. In Section II, we delineate LBP. Section III illustrates the proposed method. Section IV defines new constructive components of the genetic algorithm. Section V articulates differences between the proposed method and comparable methods. In Section VI, experiments and evaluation protocols are stated. Ultimately, we conclude the paper in Section VII.