The teaching in the Department of Computer Science Prof. Hana Mohsen Ahmed and with the participation of Hala Hassan / Baghdad University / Center for Calculation published her research entitled "The effect of levels in the transformation of the double wavelet tree when designing a comprehensive picture analyzer" in the Iraqi Journal of Science issued by Baghdad University / College of Science. The design of global image analyzers that contain implicit information is an important issue relative to the large dimensions of the characteristics of the natural image and deep neural networks, especially the deep neural networks, are ways that have an interest in the problem of designing the global image analyzer This study describes the effect of setting appropriate values ​​for a number of levels when performing pre-treatment on The image using the double wavelet transformer complex for the tree. This value has an effect on the accuracy of the detection of the images obtained to evaluate the system at the number of levels when it is equal to five gives a better performance of the system compared to the level number Other features in addition to reducing the time required for training and selecting pictures.