The measurement of the diameter of the parison is achieved by the detection of the edge of the contour image of the parison. Laptop battery edge detection are two basic approaches: one is the use of edge detection operator edge extraction, another edge extraction using gray level threshold segmentation method. From the Roberts operator, Sobel operator, LoG operator, Acer um08a31 battery Canny operator and gray threshold segmentation method to get the test results shows that the efficiency of the algorithm of grayscale threshold segmentation method detection results, the most suitable in the parison profile edge detection.

From the parison profile edge of the binary image obtaining diameter of the parison is very easy to implement. Assuming that the extruded parison its axis is perpendicular to the horizontal, the diameter of a short parison can be seen as the distance between the two points of intersection of the horizontal straight line with the contour edge. The realization method is:

1) by the line scan edge binary image, if the row pixel grayscale values ??are all 0, detects the next line, the next step otherwise.

2) If it is detected to a pixel gray scale value of 255, it is determined whether the gray scale value of the right pixel 255. The laptop battery skip the line is not detected, continue to test the next line; Otherwise, the next step.

3) to the coordinates of the point of the left edge has been detected, continue the line of the detection, scanning the right edge point. If detected, calculates the distance of two edge points Q and save continues to detect the next line; otherwise directly continues to detect the next row.

Figure 3 is on the edge of the binary image, a diameter of the detected result, the point of detection intervals of 10 pixels and the length of the parison is about 630 pixels. Wherein Figure 3 (a) is a result of the processing the gradation threshold segmentation method obtained edge contour, and Figure 3 (b) is a result of Canny operator, the processing of the detected edge contour. Diameter distribution obtained by the two algorithms are basically the same.

4 Conclusion

Directly through the camera shooting parison contour image using digital image processing technology acquisition to image processing and analysis, real-time extraction of geometric features, namely parison diameter distribution, the parison diameter can achieve real-time online testing. Laptop batteries, but the accuracy of the measurement will be affected by the influence of a number of conditions, Apple a1057 battery resolution image capture devices, lighting environments and interference. In this paper, the parison diameter distribution line detection problem discussed, it will be possible for extrusion blow molding process quality real-time monitoring and closed-loop control.