The fast image recognition based on grayscale features

The gray image is generated according to the real-time depth value on transporter in underwater terrain auxiliary navigation. The accurate position is obtained by matching to the water depth map stored on the computer of the carrier. Thus, the image recognition is an important problem. An advanced matching method based on related pixel grayscale is put forward, which combines the rapidity of the sequential similarity methods (SSDA) and the high precision advantage of normalized product correlation image matching algorithm. Experimental results demonstrate the own advantage and availability, greatly decreasing the matching time.