The segmentation of the retinal vascular tree (RVT) is undoubtedly crucial for visualizing and analyzing vessel morphology, which facilitates the detection and diagnosis of various pathologies that affect the retina. Therefore, segmentation must have high performance with reduced complexity and execution time to meet constraints imposed by the clinical context. In this context, several classic Deep Learning (DL) architectures was proposed for the automatic segmentation of the retinal vascular tree, they generate suboptimal segmentation performance and significant computational complexity. In this work, we propose a new method for RVT segmentation. The main contribution consists of suggesting a new DCNN, inspired from Segnet architecture where standard convolution blocks have been substituted by MobileNet convolution block. The proposed method enhance segmentation performance and reduce the computational complexity. It is tested on DRIVE database, and scored an Accuracy of 97.08 %, a sensitivity of 82.38 % and a specificity of 98.55%, within an architecture of 3.4 M parameters.