Traditionally, farmers and consumers have determined the ripeness of blueberries by manual means, such as observing the color, pores, and skin of blueberry fruits. Such recognition takes a relatively long time and gives rise to different levels of maturity because people have visual limitations in recognition, fatigue levels and differences of opinion about good maturity. Consumers tend to pay attention to aspects such as the striking color and size of blueberries, but do not know how ripe and nutritious the fruit is for consumption. Several image processing techniques were used in this study, including image segmentation for segmentation based on color features, expansion and contraction operations to remove noise, and naming fruit objects using recursive component labeling methods. This is followed by separation and training of geometric features and colors. At the time of testing, a classification of the degree of maturity and type of fruit of the object is carried out. To more accurately identify the degree of ripeness of the fruit, check the geometric features and characteristic values of the color content. Range values are determined by statistical methods such as mean and standard deviation. Using the K-Means algorithm to segment blueberry imagery, the study aimed to develop an efficient method to distinguish blueberry ripeness levels automatically. It can help classify and group blueberries according to their degree of maturity. In addition, the results of this study can also be used for quality control of blueberries in the agricultural industry or for fruit marketing applications.