Soil health is an aspect in crop growing to consider because extensive use of chemical inputs which most likely have left the soil very acidic. Acidic soils decreases commodity production yield. Soil analysis is a valuable tool in determining the fertility status of the soil. This study aims to develop an Inception-V3-based image classifier on pH level and apply the classier to report soil health status of rice farms automatically. Specifically, it achieved the following objectives: 1) develop a soil pH level image classifier using Transfer Learning; and 2) predict the soil health status of the rice farms using pH level image classifier. The mode of Learning is Transfer Learning via Tensorflow. The study found out that: 1) the developed soil pH level image classifier using transfer learning facilitates the production of a software with has two steps to classify pH level images; and 2) the soil health status of the rice farms in ASIST are generally has pH5.8 level. Based on the findings, we can say that soil test result can be identified easily and consistently using the pH level image classifier, thus providing a more accurate in pH level reading; and) ASIST rice farms are below the neutral level but not acidic either therefore rice farms are generally healthy. Farms are suitable for rice growing.