Abstract: Camera sorters are used both for food and non-food applications like french fries, shrimps, vegetables, snacks, raw cotton, plastics, etc. But invisible foreign bodies, such as glass, or defects with exactly the same color as the product may not be detected. So the laser scanning system is used for detecting defects when the products and the defects have the same color. A laser sorting system is designed and improved based on the selective absorption principle. And a high-speed signal sampling experiment and a imaging experiment for the system were tested. By the selective absorption sorting technology, both products and foreign bodies were identified efficiently in the signal sampling experiment. The headline in the newspaper was selected as the imaging experimental sample and the diameter of scanning spot was adjusted close to 0.35 mm. The scanning data recorded through computer program was imaged by software. The resolution of the experimental result is 180*43 pixel. Imaging results illustrate that defects (1.5 mm) can be recognized accurately. Finally, after optical scanning system was analyzed, the phenomenon was found that most of the flux was lost due to the non-perpendicular illumination of the sample. A curve scanning sorting structure was proposed for improving traditional sorting system.