The work aims to improve the edge detection accuracy of aluminum-plastic packaging materials, and design an edge detection method based on image recognition technology. The Canny edge detection algorithm was introduced. As for such problems as difficult double threshold choice and the low edge detection accuracy, high and low thresholds were selected based on Ostu algorithm. An image detection system was built based on FPGA (field-programmable gate array) and DSP (digital signal processor), and the system block diagram and the FPGA internal structure were given, namely asynchronous FIFO module, serial-in parallel-out module, pixel window refresh module, data processing module and matching circuits etc. The experimental studies were carried on with aluminum-plastic bubble cap drug package as example. The experimental results showed that the improved Canny algorithm could obtain clear and detailed images and complete edges on the premise of ensuring processing speed. This method is characterized by high accuracy, stability, reliability and faster recognition speed, and it can meet the product detection requirements in the packaging process.