Today Authentication is an essential mechanism in information security. It was used to identify authorized users and safeguard systems against hackers and spies. A graphical password (GP) is a password that employs graphics rather than text to gain access to computers. The GP is still not widely used in the actual world because users are frustrated by the numerous Login phases and the selection of various determined images from prior techniques, therefore they have returned to textual passwords. The vast majority of research has failed to discover a means to increase overall security, usability, memorability, and login time. In this paper, a novel Choice-Based Graphical Password (CGP) Scheme for Web Applications is proposed. The scheme is a two-level multifactor authentication: textual and recognition. During registration, the user first registered his/her data, and then the system assigned him/her a random and unique number. The second, user chose an image from CGP's dataset or from his/her device. CGP then resized and blurred the user-selected image before encrypting and storing it in the CGP database. The attacker claimed that the image was one element, but it was a combination of five factors (user name, user number, and his/her registered image with the same name, size, and resolution), making it difficult to guess and resistant to several attacks using The Common Attack Pattern Enumeration and Classification (CAPEC). The proposed CGP scheme's performance evaluation enhanced a 36% increase in password space, a 33% increase in possible password numbers, and a 36% increase in entropy when compared to prior methods. Our CGP approach met the challenge of the password being secure, memorable, user-friendly, and time-saving.