This design mainly studies the design of vehicle number plate classification system based on MATLAB software. The system mainly includes five core parts: image acquisition, image preprocessing, number plate location, character segmentation and character recognition. The image preprocessing module of the system converts the image into a binary image which is easy to locate the license plate through the operation of image graying, image enhancement, edge extraction and binarization. It uses the edge and shape characteristics of the number plate, and combines Roberts operator edge detection, digital image, morphology and other technologies to locate the number plate. The method of character segmentation is to find the continuous block of characters in the binary number plate, and cut if the length is longer than the set threshold, so as to complete the character segmentation. Character recognition is accomplished by template matching algorithm. Each of the above function modules is realized by MATLAB software. Finally, the number plate is identified. At the same time, the problems in the design are analyzed and dealt with concretely, and better methods are sought.
Number plate recognition system includes:
- image acquisition
- image preprocessing
- number plate location
- character segmentation
- character recognition
The system is mainly composed of image processing and character recognition. Where the image processing portion includes a map Like preprocessing, edge extraction modules, number plate positioning, and segmentation modules. Character recognition part can be divided into words Image grayscale and image edge extraction.
number plate location and number plate segmentation are the key to the entire system, and its role is in grayscale after image pre-processing. Determining the specific location of the number plate in the image and segmenting a sub-image containing the number plate character from the entire image result. For the recognition of the character recognition subsystem, the accuracy of the segmentation is directly related to the entire number plate character and the recognition rate of the system.
The ultimate goal of the number plate recognition system is to identify unclear number plate photos and output a clear picture plus outputs every number and character on the number plate
- Corrosion operation
- Image clustering, fill image
- Remove the portion of the cluster with a gray value less than 2000
- Returns the dimensions of each of the 15 dimensions, stored in x, y, z
- Tic timing starts, toc ends
- Generate a zero pin for y*1
- If the myI image coordinates are (i, j), the point value is 1, that is, the background color is blue, blue plus one
- Blue pixel count
- Y-direction number plate area determination
- Temp is the maximum value of the element of the vector yellow_y, MaxY is the index of the value
- X-direction number plate area determination
- Further confirm the number plate area in the x direction
- Correction of the number plate area
- Write colored number plates to the dw file
- Reading number plate
- Convert number plate image to grayscale image
- Write a grayscale image to a file
- T is the threshold of binarization
- Binary image
- Before mean filtering
- Filter
- Create a predefined filter operator, average is mean filtering, template size is 3*3
- D, ie, mean filtering, h for h using the specified filter h
- Some images operate
- Expansion or corrosion
- Unit matrix
- Return information matrix
- Calculate whether the ratio of the total area of the object in the binary image to the entire area is greater than 0.365
- Corrosion if greater than 0.365
- Calculate whether the ratio of the total area of the object in the binary image to the entire area is less than 0.235
- If it is smaller, the expansion operation is implemented.
- Find a block with continuous text. If the length is greater than a certain threshold, the block is considered to have two characters and needs to be split.
- Cut out 7 characters
- Split the second to seventh characters
- Normalized size in commercial system programs is 40*20
- Final output.