This work presents a motion planning approach for tomato harvesting manipulators with seven degrees of freedom (7 DOF) based on an optimization technique and alternative method. It is to find optimal joint perturbations during the path planning so that a manipulator reaches a goal from an initial position with high accuracy. The optimization model consists of the objective function defined by the tracking error and the representation of a set of mathematical relationships that describe the kinematic restrictions of the manipulator. In this method, only a forward kinematics is used and the complex mathematics in numerical solutions of an inverse kinematics is avoided to reduce the computation load. Simulation results show that the tomato harvesting manipulator can move the end-effector to the target from an initial position along a specified geometric trajectory in its workspace. Simultaneously, the joint displacements vary smoothly within their limits during tracking. The position absolute error, moving velocity and precision of the end-effector are 0.53mm, 0.18m/s and 3.75% respectively, which fulfill the requirements of tomato picking with well working efficiency.