This paper proposes a new type-2 fuzzy algorithm for the Unmanned Aerial Vehicle (UAV) image segmentation. In this algorithm, a logarithmic function is applied to transforming the membership of each pixel which is derived from the Fuzzy C-Means (FCM) algorithm; In other words, this logarithmic function defines a new membership function which can update the cluster centers effectively. The proposed method is called the Logarithmic Functional Fuzzy C-Means Algorithm (LF-FCM) used for the UAV image segmentation. The proposed method achieves better segmentation results than those of the K-means, FCM clustering algorithms and possibilistic c-means algorithm (PCM) in our comparative study for UAV images. A numerical example is provided to exam the behavior and effectiveness of all the algorithms tested. Experimental results demonstrate that the proposed LF-FCM algorithm is efficient and robust for natural UAV images. The Davies-Bouldin (DB) index, the rate of misclassification (Perror) and the