The manual revenue collection process currently operating in Tanzania, has deprived the Local Government Authority (LGA) of its potential income. Meanwhile, The Number Recognition (NR), is a modern technique that uses optical character recognition on images to obtain desired characters. The technique involves Image acquisition, pre-processing, character segmentation and Character recognition. This technique has been used in activities like license plate detection as Automatic Number Plate Recognition (ANPR). This research attempts to solve the problem by automating the current revenue collection process at the mini bus terminals in Tanzania taking a case of Dodoma Municipal using Number recognition techniques and Optical Character Recognition (OCR). The research adopted a case study research design and simulation. In addition, the study used key informant interviews and observations to acquire a good understanding of the current operations at the mini bus terminals and the necessary requirements to achieve the main goal of the study. Simulation of the Number Recognition System (NRS) was achieved by using Matlab R2017a as a simulation tool on a Dell computer running windows 7 professional, Hard Disk Drive (HDD) 500 Gigabyte (GB), Random Access Memory (RAM) 4 Gigabyte. The research, achieved a simulation for NRS with an accuracy of 0.988 for the character recognition of the captured Surface on mini buses. The researcher recommends a further study on the image acquisition process and messaging alert system, to completely automate the process.