Search

This paper studies a new color image edge detection method. The fuzzy inference engine is designed to detect edges. Some membership functions in horizontal and vertical directions are used for edge detecting. Experimental results show the presented method provides good performance when it is compared with conventional methods.

This paper studies spatial domain upsampling method. The purpose of the method is to perform high-quality upsampled results while consuming reasonable computational complexity. Experimental results prove that the proposed approach can outperform conventional methods.

This paper presents a multi-sensor color image interpretation method. Objective and subjective performance comparisons are conducted in simulation results section.

This paper presents a new circular effect generation approach. A pixel location in an image can be stated with i and j coordinate. To handle (i,j) coordinate system easier, we transform (i,j) coordinate system into (ρ,θ) coordinate system. Simulation results introduce performance comparison.

In this paper, we present a tile effect generation algorithm. The presented method cuts a given pictures with small-size tiles and an original image is composed of those tiles. The size of tiles is generated by random function. We use four parameters: size of horizontal and vertical block, size of skip blocks in horizontal and vertical direction. The presented method is tested on McM dataset. Simulation results provide performance comparison in terms of PSNR, CMSE and FSIM.

In this paper, we present a method which gives oil painting effect for color images. In general, oil painting method is a process of painting with pigments that are bound with drying oil. The oil painting is widely adopted for artist who utilizes different oils. The proposed method uses digital painting approach and makes an original image painted with oil. The main purpose of oil painting is to generate visually friendly image. The conventional oil painting method requires large computations and long execution time. However our proposed method needs less complexity due to its simplicity. Performance comparison is provided that proves the reliability of our proposed method.