The HTP test is today‘s representative projective drawing test. However, it lacks standardized interpretation criteria, and there is a high level of subjective intervention by evaluators during the evaluation process, resulting in low reliability. In order to resolve these problems, the researchers proposed an AI system HTP Object Detection Model that classified house, tree, person objects in HTP test results based on AI object detection models and sent meaningful interpretations by receiving location and size values used for formal analysis. For the study, image data of 98 houses, 98 trees, and 196 people were collected between September and October 2022. The model was trained using machine learning. A high-accuracy model was developed with an Accuracy of 90% and F1-Score of 95%. The model calculated the location and size information of the objects and provided formal interpretations that matched the corresponding values. The validity of the model was measured by art therapy experts evaluating the formal interpretation derived from HTP test images inputted into the model. The average validity was 79%. By adding diverse training data and more detailed interpretations in the future, it was expected that an formal analysis model of HTP test would be developed to supplement the projective drawing test reliability of art therapists. Based on these research findings, the significance and limitations of this study were discussed.