Deep convolutional neural networks have revolutionized many computer vision tasks, including image retrieval. The most crucial factor for using convolutional neural networks in image retrieval is their ability to extract highly representative features from images. Among existing deep networks, residual networks demonstrate higher performance in the task of image retrieval. In this paper, a new residual block is proposed to generate rich sets of edge-assisted features for image retrieval. The proposed residual block comprises three modules: the edge feature extraction module, the hierarchical feature extraction module, and the feature fusion module. Fusing the conventional features provided by the hierarchical feature extraction module and those of the features obtained from the edge feature extraction module enables the network to learn very rich feature sets. This, in turn, increases the representational ability of the obtained feature sets for image retrieval. The deep image retrieval network utilizing the proposed block is evaluated on various benchmark datasets. Experimental results confirm that utilization of the proposed residual block in a deep image retrieval network exhibits superior retrieval performance to that provided by other image retrieval networks.