Research On Automatic Recognition Algorithm Of Pointer Meter Reading Based On Image Processing

Pointer meter is a more commonly used measuring instrument.It is not only simple in structure,low in price,convenient in installation,and high in reliability,but also can intuitively display the changes of the data to be measured.Therefore,it is still widely used in the fields of industrial production,military defense,and other fields.The commonly used pointer instruments include thermometers,pressure gauges,humidity meters,flow meters,and the like.Pointer instruments usually require manual readings to obtain data.This type of reading has significant limitations: 1)The data from manual readings must be manually entered into the computer for storage and transmission;2)Accuracy and accuracy of readings are dependent The subjective awareness of the reading staff;3)The reading personnel may be placed in a high-risk,toxic environment.Therefore,it is necessary to overcome the limitations of manual readings and use an automatic recognition method to achieve pointer meter readings.This paper proposes an automatic readout recognition algorithm based on image processing.The specific algorithm is as follows:(1)Obtain image of the instrument through an industrial camera or handheld terminal device,perform pre-processing on the image,detect the possible dial outline through the contour tracking algorithm,and then determine the position of the dial area by the shape feature.(2)In the dial area,use the Line Segment Detector(LSD)algorithm to detect the straight line segment,and use the least square method to fit the circle where the midpoint of the straight line segment is located to find the center O and radius R of the dial.The tick line is screened by the distance from the midpoint of the line segment to the center of the circle.The size of the scale zone is determined by the line length and radius.In the area of tick marks,calculate the angle between each black pixel point to the center O and the X-axis and round it.Because the main tick area contains more black pixels than subdivided tick marks,the center tick method is used to extract the main tick marks.Hover transform is used to detect the edge lines on both sides of the dial in the position of the dial area,and the intersection of the edge lines on both sides is recorded as the pointer endpoint P.P is connected to the circle center O and marked as a pointer line segment.(3)Find the character positions of the main tick marks on the left and right sides of the pointer in the area around the tick marks,and obtain the readings of the main tick marks on the left and right sides of the pointer through the digit recognition algorithm.Calculate the distance from the main scale line to the left and right sides of the pointer,and then calculate the pointer number X based on the distance method.The main innovations of this algorithm are as follows:(1)Using the LSD algorithm to accurately detect the tick marks.(2)The Hough transform is used to obtain the positions of the main tick marks on the left and right sides of the pointer.The intersection points of the edge lines on both sides are recorded as the end points of the pointer,and the end points of the pointer are connected with the center of the circle as the pointer lines.(3)The existing method has known conditions such as range,zero scale,etc.The algorithm uses no prior knowledge to read the readings.The pointer meter automatic reading algorithm proposed in this article can accurately and quickly locate the scale line and scale number,and can accurately complete the one-to-one correspondence between the scale line and the scale number.The algorithm improves on the basis of the Hough transform to detect pointer segments and improves the accuracy of pointer positioning.Based on the support of OpenCV library and boost library,this algorithm validates the feasibility of the algorithm with Visual Studio development platform.Using the MFC module creation dialog box in the Microsoft Visual Studio 2013 integrated development environment,the step-by-step meter auto-reading algorithm is gradually operated by adding different buttons.This method not only has high robustness,but also can accurately and quickly identify pointer indications.