Adaptive Thresholding Algorithms and Morphological to Improve the Quality of Takepan Sasak Image Readability

Abstract

One of the cultural heritages in West Nusa Tenggara, especially Lombok, is takepan (lontar ancient manuscripts). Takepan written on leaf sheets of Lontar is known as the jejawan script (derived from ancient Javanese). local adaptive[1] adaptively setting thresholds for each pixel can behave better than the global method in degraded and damaged document images. Adaptive binarasi can also reduce noise when segmenting text from the background. The process of determining the threshold value works on blocks in the image. Images that have contrast and lighting variations that are very difficult to solve as background or background because there are a lot of pixels, this recommends using Local Adaptive Thresholding. From the research that has been done, it can be concluded that adaptive thresholding and morphology with dilation approach help maximize image readability with the highest PSNR value reaching 34,107 dB. In further research can be developed using feature extraction so that it can recognize the character of the letter in takepan.