研究目的
Developing an automated system for detecting potato defects using infrared imaging to ensure food quality and safety.
研究成果
The developed algorithmic and software tools can detect potato defects with up to 95% probability under optimal conditions. Future work includes developing detection algorithms based on neural networks to improve reliability.
研究不足
The detection probability decreases with increased transportation and rotation speeds of tubers. The temperature difference between the tuber and the rollers significantly impacts detection results.
1：Experimental Design and Method Selection:
The study uses a computer vision system to recognize potato defects in real time based on temperature differences between healthy and damaged tissues after short-term heating.
2：Sample Selection and Data Sources:
Potato tubers are used as control objects.
3：List of Experimental Equipment and Materials:
Includes a chain conveyor, infrared heaters, a Flir Ax5 thermal imager, and a personal computer for image processing.
4：Experimental Procedures and Operational Workflow:
Potatoes are fed to a conveyor, heated, and their images are captured and processed to detect defects.
5：Data Analysis Methods:
Image processing involves logarithmic correction of contrast and brightness, threshold brightness operation, and contour detection using Non-PreWitt and Sobel filters.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
