This study is an experimental attempt to construct a comprehensive archive of signature elements of Korean modern painters, Park Sugeun and Lee Jungseop, through image-based-analysis of their artworks in order to provide scientific criteria for connoisseurship. By ‘signature,’ we mean the micro- stylistic details of an artwork, which are so distinctive that a systematic analysis of them would articulate the formal characteristic of an artist, and
ultimately, make contribution to identify its creator. We particularly applied 1) Comparing Median Wavelet Coefficients at the Finest Scale 2) Comparing Gabor Wavelet Energy 3) Comparing Fourier Descriptors. Authentication of artwork has long been reliant upon the connoisseur’s expertise, yet the traditional connoisseurship is criticized as being based on intuitive judgment and empirical evidences, which are mainly informed by ‘experienced perception’ and ‘refined feeling.’ Yet, due to its overt reliance on such subjective criteria, authority of connoisseurship has come under fierce attack from those who condemn the whole enterprise as bogus. Contemporary connoisseurship is thus aided by scientific lab tests such as pigment analysis, X-ray examination, dendrochronology, infrared photographs, and etc. However, the current scientific methods mostly serve to confirm dates and reveal forgeries, but they cannot identify the artists and the works’ authenticity. The material sampling required for lab testing is damaging the works, and is not always available. Our study is to combine the traditional connoisseurship with image processing technology, in order to standardize the aesthetic judgment by making up for the limitations of both conventional art appraisal and technology-oriented process. This project is a highly advanced customization of the Morellian method of connoisseurship based on an accurate and systematic perception and on the comparison of details. It is an experiment, in a sense that it can be understood both as a reinterpretation of the existing scientific method and as a hypothetic method whose basis should be further justified by empiricism. By archiving the repetitive visual patterns of the modern Korean masters, our team aimed to define the elements that most clearly manifest the artist’s individuality; develop mathematical analysis of the elements; create a scientific reservoir of the individual artist’s signature elements. In this way, such empirical criteria as ‘perception,’ ‘feeling,’ and ‘intuition’ will be mathematically systemized. This quantitative analysis of artwork will possibly respond to the art-historical and technical questions to identify the nature of artistic appreciation, but it is by no means a key to all questions in the field of connoisseurship.