A realtime detection and processing system for contaminants on chicken carcass surface was designed based on machine vision technology. The surface feature images in the wavelengths 500nm and 710nm of chicken carcass were acquired by means of industrial camera, filter and computer respectively. These images were firstly processed with the median filtering and gray enhancement. Then the binarized images were obtained by using Otsu automatic thresholding. Then, the corrosion, expansion and void filling operations were conducted on these images, and the contamination areas on chicken carcass surface were segmented. On the basis of this, if contaminations (cecal fecal, blood, bile) on the surface of chicken carcass were distinguished, the spray treatments for these contaminations were carried out. The experimental results showed that the total average correct ratio was 90.5% when this system was used to detect three contaminations on the surface of chicken carcass, implying that this system could be used for online detection of contaminants on chicken carcass surface with high detection accuracy and reliability.