This paper studies an approach for estimating the mass and volume of Vietnam-ese avocado 034 using top-view image sample combined with geometric slice modeling. The method involves analyzing image-based geometric features to per-form efficient mass prediction. Key components are defined and collected from top-view images through image processing techniques, including length, width, and height. The selected features later are integrated into the geometric slice mod-eling that approximately estimates the mass of fruit. The approach is designed to support agriculture and post-harvest processing by offering a practical and scala-ble alternative to traditional weighing methods. A dataset of 101 avocado samples was used to evaluate the method under different slicing configurations (3, 5, 7, 8, 10, 12, 15, 18, and 20 slices). Experimental results show that the accuracy of weight estimation reach up to 91.35% by slicing 12 slices. Moreover, the pro-cessing speed of each image with configuration with 7 to 12 slices ranged from 0.148 to 0.168 seconds, which corresponds to real-time processing of up to 5–7 images per second on standard hardware. This approach is low-cost, non-destructive, and scalable, making it suitable for agricultural and post-harvest ap-plications such as automated fruit sorting and grading.