CREATION OF IMAGE MODELS FOR INSPECTING DEFECTS ON COMMERCIAL DRIED FISH FLOSS

Abstract

Both fish and fish related products are two common foods in our daily lives. Since fish products have highly economic value, the main form of direct consumption of raw fish, cooked fish, processed products (such as preserved fish, canned fish, fish floss, etc.) is fairly extensive. Fish floss is a chopped finely or mashed fish meat boiled in seasonings, then stir fry until the fish meat is arid and pulverous. When fish is usually eaten, fish bones are some of the most regularly gulped extraneous objects happened in meals. In the making of commercial fish flosses, fishbone inspection is conducted by expertise inspectors using their feeling of contact and sight which could cause misjudgments. When consumers eat fish floss with bones, it may cause harm to the health of consumers. Therefore, this study proposes an automated fishbone inspection method for fish floss products. The proposed method applies the curvelet transform with square-ring low-pass energy filtering to remove the random patterns of background and delete the angle direction of background texture. The approximated and partial detailed components regarding fish bones and uniform background are preserved in the low and medium frequency bands. Then the filtered image is inversely converted to spatial domain. In the reconstructed image, the background random texture is attenuated and the fishbone areas are enhanced. Finally, a threshold value is determined by statistical interval estimation and the restored image can be easily segmented to into two categories namely dark fish bones, and white background. Outcome results show that the suggested inspection system can effectively determine the presence and the locations of fish bones on the surfaces of fish floss.