CN102779342A - Edge detection method for voice coil motor magnet - Google Patents

Abstract

The invention provides an edge detection method for a voice coil motor (VCM in short) magnet based on mathematical morphology and a canny operator, belonging to the field of computer image identification. The edge detection method comprises the following base steps of: (1), inputting a VCM image to carry out background removal and binarization processing, so as to obtain a binary image of the magnet; (2), firstly carrying out erosion operation and expansion operation on the binary image in sequence, and then carrying out expansion operation and erosion operation in sequence; (3), using a Gaussian filter to carry out smoothing on the image acquired in the step (2); (4), carrying out gradient computation to acquire a gradient value of each pixel point; (5), carrying out non-maxima suppression on each pixel point and gradient value to acquire a preliminary edge point; and (6), carrying out double-threshold computation on the preliminary edge point acquired in the step (5), so as to acquire a complete edge. The edge detection method provided by the invention can be used to stably and efficiently detect the edge of the VCM magnet.

Description

A kind of voice coil motor magnet edge detection method

Technical field

The present invention relates to a kind of voice coil motor magnet edge detection method, belong to the computer picture recognition field, be applicable to the edge extracting of voice coil motor magnet based on mathematical morphology and canny operator.

Background technology

Voice coil motor (Voice Coil Motor is called for short VCM) is an indispensable part in the hard disk; Magnet above it can form magnetic field; Utilize this magnetic field to drive the reading-writing arm in the hard disk, so the marginal information of VCM magnet is one of key factor that guarantees the hard disk quality.Yet in actual detected,, tend to occur false edge when gathering the magnet marginal information, cause location, magnet edge inaccurate because the VCM shape is comparatively complicated.

Mathematical morphology is that the basis is analyzed image with geometry; Structural element is applied to input picture; Thereby obtain and the equal big or small output image of input picture, its basic thought is usually to survey and extract characteristics of image to reach the purpose to graphical analysis and identification with the structural elements with certain form.Expansion, burn into open and closure is computing basic in the mathematical morphology.

The Canny operator is a multistage optimization operator with filtering, enhancing and detection.Canny provides 3 indexs that evaluation edge detects performance:

1) high s/n ratio criterion, promptly non-marginal point are mistaken for marginal point or marginal point to be mistaken for the probability of non-marginal point low, make the signal to noise ratio (S/N ratio) of output reach maximum;

2) high position precision criterion, promptly detected image border point will be as far as possible at the center at real image edge;

3) single skirt response criterion will guarantee that promptly single edge has only a pixel corresponding, and the response of false edge is inhibited to the full extent.

Summary of the invention

To the problems referred to above, the object of the present invention is to provide a kind of VCM magnet edge detection method based on mathematical morphology and canny operator, it can real-time stabilization detects the edge of VCM magnet efficiently.

VCM magnet edge detection method based on mathematical morphology and canny operator may further comprise the steps:

Step 1), the tiff image of input VCM is removed background and binary conversion treatment, obtains the bianry image of magnet;

Step 2), be that 15 collar plate shape structural element carries out the mathematical morphology computing of corrosion after expansion earlier to bianry image utilization radius, promptly open computing; And then the utilization radius is that 8 diamond structure element carries out the computing of first expansion post-etching, i.e. closure operation;

Step 3) adopts Gaussian filter to step 2) image that obtains carries out smoothing processing, the filtering image noise;

Step 4) adopts the canny operator to carry out gradient calculation, obtains the Grad of each pixel in the image;

Step 5), the Grad of each pixel in the image that the employing step 4) obtains carries out non-maximum value inhibition to each pixel and Grad, obtains preliminary image border point;

Step 6), the preliminary marginal point that step 5) is obtained carries out dual threshold calculating, and the marginal point of further refining is rejected false marginal point and completion emargintion, obtains complete edge.

The present invention adopts technique scheme; Utilize the mathematical morphology algorithm that the VCM image is carried out pre-service and image characteristics extraction and can effectively remove the detailed information of disturbing and keeping image, the image utilization canny operator after handling is carried out rim detection can obtain VCM magnet shape edge clearly.

Description of drawings

Fig. 1 is the process flow diagram based on the VCM magnet edge detection method of mathematical morphology and canny operator.

The VCM original graph of Fig. 2 for gathering.

Fig. 3 is the pretreated synoptic diagram as a result of mathematical morphology.

Fig. 4 is a magnet edge detection results synoptic diagram.

Embodiment

Further specify below in conjunction with the accompanying drawing specific embodiments of the invention.

1) input needs the VCM picture of detection, does the removal background process, extracts interested magnet part, does binary conversion treatment, obtains the bianry image X of magnet.

2) choose structural element S ₁And S ₂, S wherein ₁For radius is 15 collar plate shape structure, S ₂It for radius 8 diamond structure.Utilization S ₁X is carried out the mathematical morphology computing of corrosion after expansion earlier, promptly open computing Obtain X ₁, and then utilization S ₂Carry out the mathematical morphology computing of first expansion post-etching, i.e. closure operation $X_1\·S_2=(X_1\⊕S_2)\mathrm{\Θ}{S}_2,$ Obtain X ₂

3) adopt two-dimensional Gaussian function $G(x,y)=\frac{1}{{2\mathrm{\π\; \σ}}^2}{e}^{\left(\frac{-(x^2+y^2)}{{2\mathrm{\σ}}^2}\right)}$ Try to achieve first order derivative $\frac{\∂G}{\∂x}=\frac{-x}{{2\mathrm{\π\; \σ}}^4}{e}^{(-\frac{x^2+y^2}{{2\mathrm{\σ}}^2})},$

With X ₂Carrying out convolution algorithm obtains

With

Calculate the Grad of each pixel in the image $A(i,j)=\sqrt{E_x^2(i,j)+E_y^2(i,j)}.$

4) each pixel and Grad being carried out that non-maximum value suppresses is the ridge band of refinement gradient magnitude image; Only keep the maximum point of amplitude localized variation; If pixel (i; J) (i j) is not less than the gradient magnitude of two neighbor pixels on the gradient direction to edge strength A, judges that then this point is possible marginal point.

5) set two threshold tau ₁And τ ₂, non-maximum value is suppressed image carry out the dual threashold value, can obtain two testing results, be respectively with low threshold tau ₁Cut apart the image T that obtains ₁With with high threshold τ ₂Cut apart the image T that obtains ₂With image T ₂Be basic adjoining edge profile, when being connected to end points, with image T ₁Carry out the final rim detection design sketch of edge connection acquisition for replenishing.

More than being preferred embodiment of the present invention, is not that the present invention is done any pro forma restriction, and every foundation technical spirit of the present invention all belongs in the scope of inventing technical scheme any simple modification, equivalent variations and modification that above embodiment did.

Claims (2)

1. a voice coil motor magnet edge detection method is characterized in that: adopt mathematical morphology that voice coil motor magnet image is handled earlier, use the canny operator to detect magnet edge shape clearly again.

2. require described voice coil motor magnet edge detection method according to right 1; It is characterized in that: it is to be that 15 collar plate shape structural element carries out the mathematical morphology computing of corrosion after expansion earlier to bianry image utilization radius that the mathematical morphology that is adopted is handled image, promptly opens computing; And then the utilization radius is that 8 diamond structure element carries out the computing of first expansion post-etching, i.e. closure operation.

Images