Lip Detection and Recognition-A Review1
Dr. D. Malathi,Saloni Sharma
DOI: https://doi.org/10.1109/ICNWC57852.2023.10127483
2023-04-05
Abstract:It’s no secret that security systems rely heavily on image processing because of its versatility. Two-dimensional visuals, intricate algorithms, and instantaneous decision-making are all challenges that must be met by the system. It is possible to optimize the system at one of four stages: preprocessing, feature extraction, Lip detection, and Recognition. Using modern computing hardware and software, we can create a system that is both easy to use and exactly what we need. Unfortunately, as more characteristics are added, the complexity of implementing these algorithms grows. The process is improved through the development of novel approaches, tools, and strategies. Machine learning and AI techniques have recently been applied to image processing applications. Standard methods of authentication, such as passwords and PINs, are becoming increasingly insecure. Physical and biological characteristics that are unique to each individual provide the best level of security. It is vulnerable to guessing and theft in business and public computer networks. Plastic cards, smart cards, and computer token cards all have non-security flaws in the form of forgery, loss, corruption, and inaccessibility. Identifying techniques based on biometrics have several applications in forensics, finance, and other fields. Voluntary action from the past has the drawbacks of being difficult to implement and not adaptable for covert uses, such as in surveillance applications. Lip image audit and verification during biometrics record keeping is prone to human error. Image quality of the lips is more easily obtained than fingerprint images. Only about five percent of the population has imperfect fingerprints and cannot be verified. Reasons include but are not limited to dry skin, diseased skin, elderly skin, wounded skin, calloused finger, oriental skin, bandaged finger, narrow finger, smeared sensor on reader, etc. Varying lighting conditions are widely recognized as one of the most crucial aspects for accurate Lip recognition but also one of its greatest obstacles. Simultaneously, the same person's lip expression can look extremely different depending on the illumination.
Computer Science