A car's exhaust system not only purifies exhaust gases but also reduces noise, significantly impacting a vehicle's overall performance. Therefore, efficient and high-quality assembly is crucial, requiring a high-performance monitoring system on the assembly line to guarantee product quality. Assembly Direction Issues As a vital component for exhaust emission and purification, the assembly direction of the ceramic catalyst directly affects the purification function. If the assembly direction is inconsistent with the airflow direction, purification will fail, resulting in substandard emissions. The same applies to the muffler assembly. Since most mufflers are symmetrical, manufacturers engrave airflow direction arrows on them to identify the direction. Incorrect welding direction will inevitably affect the muffler's noise reduction effect. Both of these assembly errors cause significant inconvenience to users, necessitating a high-performance monitoring system on the automotive exhaust system assembly line to ensure product quality. Currently, most exhaust system companies use two inspection methods: manual visual inspection and barcode scanning using a barcode sensor. For the two detection methods mentioned above, the first method is prone to misjudgment due to human error; the second method requires barcode affixing before detection, increasing the labor intensity of workers, and the usability of the barcodes decreases with use. OCR Character Recognition System To address this issue, Suzhou Weisheng Machine Vision Technology Co., Ltd. has developed an OCR character recognition system. This system can be used for the recognition of various workpieces, character recognition, and the identification of character printing or print quality, not only solving the risks in the assembly process but also eliminating the need for barcode affixing. 1. System Composition The system consists of an illumination source, CCD camera, lens, computer, image acquisition, DIP processing software, control unit, monitoring unit, and alarm rejection unit, as shown in Figure 1. 2. Main Functions This system directly identifies characters at both ends of the ceramic carrier through the CCD camera to distinguish the positive and negative assembly direction of the ceramic carrier. This is a fairly mature character recognition technology in current vision inspection systems, capable of accurately capturing and recognizing characters. In addition to its error-proofing function, this system also has another important function—traceability. The system can match the characters on the identified ceramic carrier with the production date and product number on the outer casing, and store the identified characters in various formats such as Excel spreadsheets using data management software (see Figure 2). This is essential for tracing product issues during future use. Because the characters on the ceramic carrier are invisible after it is sealed, the traceability function is particularly important. We can trace back to the batch number of the ceramic carrier used for a specific model of product manufactured on a specific date. Furthermore, if a worker uses the wrong batch, the system will provide an alarm message to alert the worker. The system has alarm functions (including audible and visual alarms), and when an assembly error is detected, not only will an alarm message be issued, but the machine will also be completely immobilized until the worker adjusts the product assembly until it is completely correct. In conclusion, the development of the OCR character recognition system not only achieves a perfect combination of a vision system and an automatic detection alarm and rejection system, but its user-friendly interface and ease of operation and maintenance also significantly save users production costs.