Complexity of multi-axis, multi-functional computer numeric control machine tools and high productivity requirements with high precision is increasing day by day. This enhances the growing importance of machine intelligence, automation and safety. Integration of different types of manufacturing process, for example, Mill-Turn is the need of modern technology but this integration also increases the profile of collision risks. The general context of this study is the multi-axis production process automation and intelligence (to work and react automatically) for safe tool movements. The work done during this research should appear in the form of machine intelligence for automatic safe and efficient collision-free tool trajectory generation without collisions. This study focuses on development of an intelligent approach for enhancing decision processes by providing automatic solutions to the collision problem. The main objective is to make machine vision control effective, that is, images or video processing, so that it can see activities regarding collisions and can react or command automatically for safe production. In the presence of obstacles, the proposed approach will provide decisions regarding trajectory correction and improvement. The proposed vision tool is able to take into account the evolution of the scene, such as the aspects of obstacle changes (shape, size or presence) during production. Results presented are based on automatic 2D traversal safe trajectory generation and correction in a virtual simulated but dynamic environment. The approach presented will be adapted to online collision problems on the shop-floor by integrating it with step-compliant data interface for numeric controls technology in the future.