With the passage of time, paintings can be damaged, and common deteriorations found in ancient paintings include cracking. Cracks can be caused by many factors, such as ageing, drying, and mechanical factors. The detection and restoration of crack formation on the earliest digitized paint surface concede great significance and safety for cultural heritages. particular, this paper is based on image processing methodology to detect and discriminate cracks. First, images are enhanced in the preprocessing stage for further processing. Then two proposed algorithms are employed to detect and separate medium and thin crack images. The experiments show better results compared with previous work. This study also confirms that the proposed image processing algorithms are an efficient and robust tool for the detection and discrimination of cracks.