The presentation is about the neural networks which is I implemented in the training simulator visual systems in the form of evaluation tool.The training simulation is now a days using for the various applications in the vehicle and the aircraft systems. The visual interface is also a part of the simulators which works as the simulation of the optical view and also to send the infra red, sonar and radar information. 

The graphics based simulation has some connections between the images and the ability of the system. To make effective system the designing of the system is done as required for the development of the application. The requirements are decided only after being analyzed on the humans. The project is focused on the implementation of the neural network to conclude the evaluation problem. 

The neural network use in the system provides the task oriented calculation of an algorithm which leads to make an exact algorithm for the development of the application. For the algorithm testing, various algorithms are applied and evaluated. The neural network is expertise and analyzed for every work and the algorithm. 

The neural network consists of the three layers network that extends the feed to make the competitive learning and the controlled back propagation training. 

The project analyzed in the field of the resolution and sampling, shading algorithms, texture anti- aliasing technique and the machine complexity.

The realistic simulator application one can think of the cutoff points that will be different from every object and every task. The neural network technique is able to identify the cutoff points and that would not use lot of time to explore the large number of polygons. 

The Phong shading reduces and inhibits this cutoff and explore only small number of polygons to identify and easy to implement.