ﻃﺮاﺣﯽ و اﺟﺮاي ﺳﯿﺴﺘﻢ ﺗﺸﺨﯿﺺ ﺧﻮدﮐﺎر ﻧﺎﺣﯿﻪ ﭘﻼك ﺧﻮدرو ﺑﺮاي ﮔﯿﺖ ﻫﺎي ورودي اﻣﺎﮐﻦ ﺣﻔﺎﻇﺘﯽ

Design and Implementation of Automatic License Plate Recognition System for Security Gates

در ﺳﺎلﻫﺎي اﺧﯿﺮ، ﮐﻨﺘﺮل ﻫﻮﺷﻤﻨﺪ ﻣﺒﺎدي ورودي و ﺧﺮوﺟﯽ، ﺑﻪوﯾﮋه ﻣﺒﺘﻨﯽ ﺑﺮ ﭘﺮدازش ﺗﺼﺎو ﯾﺮ، ﺑﺴﯿﺎر ﺗﻮﺳﻌﻪ ﯾﺎﻓﺘﻪ اﺳﺖ. اﻃﻼﻋﺎت ﺑﻪدﺳﺖآﻣﺪه از ﭼﻨﯿﻦ ﺳﯿﺴﺘﻢﻫﺎﯾﯽ ﺑﺮاي ﻧﻈﺎرت ﺑﺮ ﺗﺮاﻓﯿﮏ، ﻋﺒﻮر و ﻣﺮور وﺳﺎﯾﻞ ﻧﻘﻠ ﯿﻪ ﻣﻬﻢ ﻣﯽﺑﺎﺷﺪ. اﯾﻦ روﯾﮑﺮد ﻣﯽﺗﻮاﻧﺪ در ﺣﻔﻆ اﻣﻨﯿﺖ ﻋﻤﻮﻣﯽ ﻧﻘﺶ ﻣﺆﺛﺮي داﺷﺘﻪ ﺑﺎﺷﺪ و در ﺑﻪ دﺳﺖ آوردن آﻣﺎر ﺗﺮاﻓﯿﮏ ﻣﺎﻧﻨﺪ ﺷﻤﺎرش ﺗﺮدد در دروازه ﻫﺎي ﺷﻬﺮﻫﺎ، ﺑﺮآورد آﻣﺎري ﻣﺸﺘﺮﯾﺎن از ﺳﺎزﻣﺎن ﻫﺎ، ﺷﻨﺎﺳﺎﯾﯽ و ردﯾ ﺎﺑﯽ ﻋﺒﻮر و ﻣﺮور ﻣﺸﮑﻮك ﻧﻘﺶ ﻣﺆﺛﺮي داﺷﺘﻪ ﺑﺎﺷﺪ. وﺳﺎﯾﻞ ﻧﻘﻠﯿﻪاي ﮐﻪ وارد ﯾ ﮏ ﺳﺎزﻣﺎن ﭘﺮﺗﺮدد ﻣﯽ ﺷﻮﻧﺪ، در زﻣﺎن ورود و ﺧﺮوج ﺷﻨﺎﺳﺎﯾﯽ ﺷﺪه و ﮔﺰارش ﺗﺮاﻓﯿﮏ آنﻫﺎ، از ﺟﻤﻠﻪ ﺗﺼﻮﯾﺮ وﺳﯿ ﻠﻪ ﻧﻘﻠﯿ ﻪ ﺑﺎ ﺷﻤﺎره ﭘﻼك و ﺗﺎرﯾﺦ و زﻣﺎن ﺗﺮدد، ﺛﺒﺖ ﻣﯽﺷﻮد. در اﯾﻦ ﺗﺤﻘﯿﻖ، ﻣﺎﯾﮏ روش ﺟﺪﯾﺪ ﺑﺮاي اﺳﺘﺨﺮاج ﻧﺎﺣﯿﻪ ﭘﻼك ﺧﻮدرو ﭘﯿﺸﻨﻬﺎد ﻣﯽدﻫﯿﻢ. روش ﻣﺎ از ﺗﺮﮐﯿﺐ روش ﺗﺸﺨﯿﺺ ﻟﺒﻪ ﻋﻤﻮدي Canny و ﺑﺮرﺳﯽ اﺟﺰاي ﻣﺘﺼﻞ ﺑﺮاي ﺷﻨﺎﺳﺎﯾﯽ ﻣﻨﺎﻃﻖ ﻧﺎﻣﺰد ﭘﻼك اﺳﺘﻔﺎده ﻣﯽﮐﻨﺪ. آزﻣﺎﯾ ﺶ در ﺷﺮاﯾﻂ واﻗﻌﯽ )ﺗﺼﺎو ﯾﺮ ﻣﺴﺘﺨﺮج از دورﺑﯿﻦﻫﺎي ﻧﻈﺎرﺗﯽ در ﺳﺎزﻣﺎن ﺣﺎﻣﯽ اﯾﻦ ﺗﺤﻘﯿ ﻖ( ﻧﺸﺎن ﻣﯽدﻫﺪ ﮐﻪ روش ﭘﯿﺸﻨﻬﺎدي ﻣﯽﺗﻮاﻧﺪ ﻧﺎﺣﯿﻪ ﭘﻼك ﺧﻮدروﻫ ﺎي ﻣﺨﺘﻠﻒ را در ﺷﺮاﯾﻂ ﻧﻮري ﻣﺨﺘﻠﻒ ﺑﺎ دﻗﺖ و ﺧﻮاﻧﺎﯾﯽ ﺑﯿﺶ از 98درﺻﺪ ﺷﻨﺎﺳﺎﯾﯽ ﮐﻨﺪ و ﻋﻤﻠﮑﺮد ﺑﻬﺘﺮي در ﻣﻘﺎﯾﺴﻪ ﺑﺎ روشﻫﺎي ﺟﺪﯾﺪ ﻧﻤﺎﯾﺶ دﻫﺪ.

In recent years, intelligent control of entry/exit points particularly based on image processing extracted from surveillance cameras, has been developed. The information obtained from such systems is important for monitoring the traffic and passage of vehicles. It can play an effective role in maintaining public safety and obtaining traffic statistics such as counting traffic at the gates of cities, statistical estimation of clients to organizations, and detecting and tracking suspicious traffic. Vehicles that enter a busy organization are identified at the time of entry and exit and a traffic log including the vehicle image with the license plate number and the date and time of entry and exit are recorded. In this research, a new method for extracting the license plate area is presented. The proposed method uses a combination of Canny vertical edge detection method and examination of connected components to identify candidate areas for the license plate. Experiment results in real conditions (images were taken from surveillance cameras in the organization supporting this research) showed that the proposed method can identify the license plate area of different vehicles in different lighting conditions with accuracy and readability of more than 98% and presents better performance than other state of the art methods.