Sea-sky-line detection method based on Gaussian mixture model and texture analysis

• Main Authors: LI HENGYU, LIU JINGYI, LUO JUN, YANG WANHUI, XIE SHAORONG
• Format: Patent
• Language: Chinese; English
• Published: 01.10.2019
• Subjects: CALCULATING; COMPUTING; COUNTING; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; PHYSICS

The invention discloses a sea-sky-line detection method based on a Gaussian mixture model and texture analysis. The sea-sky-line detection method comprises the following steps: (1) inputting a sea surface image to be detected; (2) dividing the color sea surface image to be detected into a plurality of uniform areas; (3) roughly segmenting the image according to texture analysis to obtain pixel classification prior information; (4) initializing Gaussian mixture model parameters through priori information of the pixels; (5) maximizing the posterior probability of pixel classification by using anEM algorithm, and obtaining a fine-grained segmented image; and (6) extracting a boundary line of the seawater area as a detected sea antenna. According to the method, the sea surface image probability distribution model is obtained by using the sea surface image texture information and the Gaussian mixture probability model, finally, the sea surface area boundary line is extracted as a sea antenna detection result, and