Objective: To solve the problems of low efficiency and poor motion stability of DELTA robots in the tomato sorting process of existing tomato processing production lines. Methods: Based on the analysis of the DELTA robot sorting system, the optimized gate shaped trajectory of the cyclotron was used for DELTA robot trajectory planning. By replacing the right angle part of the gate shaped trajectory with a convolutional line, an optimization model with the goal of time optimization was established. The optimized parameters of the gate shaped trajectory were solved using an improved ant colony optimization algorithm. Analyzed its performance through experiments. Results: The proposed method had better average sorting time and sorting success rate compared to the gate shaped trajectory and other methods, with an average sorting time reduction of 11.61% and 1.63%, respectively, and a sorting success rate increase of 2.05% and 0.56%. Conclusion: The proposed planning method effectively improves the sorting accuracy and efficiency of DELTA robots.