JPH09189529A - Shape recognition method and apparatus using image obtained by illuminating shape recognition target with shape illumination - Google Patents

Abstract

In a conventional shape recognition device, when recognizing a shape of a shape recognition target object in which a change in a sphere, a cone, and unevenness is small,
In order to perform shape recognition using images of the shape recognition object from a plurality of positions, it is necessary to provide a mechanism for moving one camera and a plurality of cameras. For this reason, the shape recognition device becomes large in size and expensive. SOLUTION: A shape recognition target object is irradiated with arbitrary shape illumination, and image data of the irradiated shape recognition target object is captured,
The shape of the shape recognition target is recognized by comparing the captured image data with the image data of the image data library that stores the image data as the shape determination reference. This allows
Since shape recognition can be performed with one image, it is not necessary to provide a camera moving mechanism or a plurality of cameras, and the shape recognition apparatus can be downsized and inexpensive.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a shape recognition method and apparatus using image processing, and more particularly to a shape recognition method and apparatus using an image obtained by irradiating a shape recognition object with arbitrary shape illumination. is there. By this method, the shape recognition device using image processing is downsized and reduced in cost.

[0002]

2. Description of the Related Art Image processing technology for recognizing the shape of an object is widely used as a means for identifying and inspecting the object. However, when recognizing the shape in the perspective direction of the shape recognition target object from the camera facing the shape recognition target object, for example, a shape recognition target object with a small change in spheres, cones, and irregularities has a change in brightness and darkness under normal illumination. Shape recognition is difficult because it is small. Therefore, an image of a shape recognition target object can be taken from multiple positions by moving one camera, or can be taken from multiple positions by installing multiple cameras and a shape using multiple images can be obtained. We are aware.

FIG. 11 shows a configuration diagram of a conventional shape recognition apparatus provided with two image capturing devices (for example, cameras). In the figure, 111 is a shape recognition target, 112 is a lighting device that illuminates the shape recognition target 111, 113 is an image capturing device that captures image data of the shape recognition target 111 irradiated by the lighting device 112, and 114 is An image processing device that uses the image data of the shape recognition target object 111 captured by the image capturing device 113 as a binary pattern, and 115 represents the image data of the shape recognition target object 111.
The computer determines the shape by comparing the binary pattern processed in step 4 with the binary pattern of the image data as the shape determination reference.

[0004]

However, FIG.
In the conventional shape recognition device shown in, when recognizing the shape of the shape recognition target object of which the change of the sphere or the cone and the unevenness is small, the shape recognition is performed using the images of the shape recognition target object from a plurality of positions, A mechanism for moving one camera and multiple cameras must be provided. Therefore, there is a problem that the shape recognition device becomes large in size and expensive.

[0005]

The present invention has been made in consideration of the above problems, and irradiates a shape recognition target object with arbitrary shape illumination, and obtains image data of the irradiated shape recognition target object. The captured image data is compared with the image data of the image data library that stores the image data as the shape determination reference, and the shape of the shape recognition target is recognized. With this, since shape recognition can be performed by one image, it is not necessary to provide a camera moving mechanism or a plurality of cameras, and the shape recognition apparatus can be downsized and low-priced.

[0006] As used herein, the term "shape illumination" refers to illuminating an object with light having a shape (such as a circular spotlight).

[0007]

BEST MODE FOR CARRYING OUT THE INVENTION The present invention irradiates a shape recognition target object with arbitrary shape illumination, captures image data of the irradiated shape recognition target object, and captures the captured image data and the image data used as a criterion for determining the shape. The shape of the shape recognition target object is recognized by comparing the stored image data of the image data library. With this, since shape recognition can be performed by one image, it is not necessary to provide a camera moving mechanism or a plurality of cameras, and the shape recognition apparatus can be downsized and low-priced.

FIG. 1 is a block diagram showing the principle of the present invention. In the figure, 1 is a shape recognition target, 2 is a shape lighting device that illuminates the shape recognition target 1, and 3 is an image capturing device that captures image data of the shape recognition target 1 irradiated by the shape lighting device 2. Reference numeral 4 is an image processing device that uses the image data of the shape recognition target object 1 captured by the image capturing device 3 as a binarization pattern, and 5 is a binarization pattern in which the image data of the shape recognition target object 1 is processed by the image processing device 4. And a binarized pattern of image data, which is a criterion for determining the shape, and determines the shape.

FIG. 2 shows an image pickup view of an image capturing device when a shape recognition object (a disk and a sphere) is illuminated with a rectangular shape illumination. (1) is a case where a rectangular shape illumination is applied when the shape recognition target is a disk, and the shape of the irradiated part remains rectangular, so the surface of the shape recognition target is flat. Therefore, it can be determined that the shape recognition target object is a disk. (2) is a case where a rectangular shape illumination is applied when the shape recognition target is a sphere, and since the shape of the irradiated part is transformed into a fan shape, the shape recognition target may have a rounded surface. As a result, it can be determined that the shape recognition target has a spherical shape.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 3 is a block diagram showing the appearance inspection apparatus for a printed circuit board according to an embodiment of the present invention. In the figure, 301 is an inspection object, 302 is an XY table, 303 is an XY table controller, 304 is shape illumination, 305 is an illumination controller, 306 is a camera, 307 is an image processing device,
308 is a monitor and 309 is a computer.

The object 301 (printed circuit board) to be inspected is X
Place it on the Y table 302 and place the XY table 302 in XY
The table controller 303 positions it at a predetermined location. For example, a semiconductor device flat package (hereinafter referred to as QFP) placed at a predetermined position of the inspection object 301.
The leads are irradiated with the shape illumination 304. Shape illumination 304
The illumination controller 305 controls the intensity of illuminance and the time of strobe irradiation. QFP illuminated by shape illumination 304
Image data is captured by capturing an image of the lead of the above with the camera 306. The captured image data is processed by the image processing device 307.
The binary pattern is processed by. Computer 30
According to 9, Q of the good quality printed circuit board already registered
The binary pattern of the FP lead is compared with the binary pattern of the image processing device 307 for the inspection this time, the non-defective product and the defective product are judged, and the inspection result is displayed on the monitor 308. The computer 309 controls the XY table controller 303, the illumination controller 305, the image processing device 307, and the monitor 308.

FIG. 4 is a flow chart showing the operation of the embodiment of the present invention. The operation will be described according to this flow.

First, a computer 30 to be inspected by a computer
Position the XY table 302 so that the first inspection portion of 1 is located directly below the camera 306 (step S4).
01). The shape illumination 304 is applied to the inspection portion (step S402). Image data is captured by imaging the reflected light of the emitted light with the camera 306. The captured image data is converted into a binary pattern by the image processing device 307 (step S403). The binarization pattern of the inspection portion sent from the image processing device 307 is compared with the binarization pattern of the non-defective product already registered in the computer to determine the pass / fail (step S404). It is determined whether the inspection of all the inspection parts is completed (step S405). If the inspection is completed, the process proceeds to step S407. If the inspection has not been completed, the position of the next inspection portion is set (step S406). Then, the process returns to step S401. The inspection results of all the inspection parts are displayed on the monitor 308 (step S407).

Next, the criteria for determining whether the QFP leads are soldered normally or abnormally will be described.

FIG. 5 is a side view when the QFP leads are normally solder-bonded. In the figure, 51 is QFP, 5
2 is a QFP lead, 53 is shape illumination (rectangle), 54
Is solder, 55 is a footprint, 56 is a printed circuit board, and 57 is an XY table.

FIG. 6 shows a read image diagram when a shape illumination (rectangle) is applied to the solder surface when the QFP leads are normally solder-bonded. In the figure, 61 is QFP,
62 is a QFP lead, 63 is a read image, and 65 is a footprint. The fan-shaped central portion of the read image 63 is a characteristic of normal solder bonding.

FIG. 7 is a side view when the QFP leads are not soldered. Further, FIG. 8 is a read image diagram when the shape surface (rectangular) is irradiated on the solder surface when the QFP leads are not soldered. Reference numeral 83 in FIG. 8 is a read image, and other than that is the same as in FIGS.

The read image 83 in FIG. 8 is rectangular,
It can be judged as abnormal because the central portion is not fan-shaped.

FIG. 9 is a side view when the QFP leads are lifted. Further, FIG. 10 is a read image view when the shape surface (rectangular shape) is applied to the solder surface when the QFP lead floats. Reference numeral 103 in FIG. 10 is a read image, and other than that is the same as in FIGS. 5 and 6, and description thereof will be omitted.

The read image 103 in FIG. 10 can be determined to be abnormal because the center portion on the right side is not fan-shaped.

[0021]

The present invention is carried out in the form as described above and has the following effects.

The shape recognition target object is irradiated with arbitrary shape illumination, and the image data of the irradiated shape recognition target object is fetched,
The shape of the shape recognition target is recognized by comparing the captured image data with the image data of the image data library that stores the image data as the shape determination reference. This allows
Since shape recognition can be performed with one image, it is not necessary to provide a camera moving mechanism or a plurality of cameras, and the shape recognition apparatus can be downsized and inexpensive.

[Brief description of the drawings]

FIG. 1 is a principle configurational diagram of the present invention.

FIG. 2 is an image pickup diagram of an image capturing device when a shape recognition target (a disk and a sphere) is illuminated with rectangular shape illumination.

FIG. 3 is a configuration diagram of an appearance inspection apparatus for a printed circuit board that is an embodiment of the present invention.

FIG. 4 is a flowchart showing the operation of the embodiment of the present invention.

FIG. 5 is a side view when the QFP leads are normally solder-bonded.

FIG. 6 is a read image diagram when a shape illumination (rectangle) is applied to the solder surface when the QFP leads are normally solder-bonded.

FIG. 7 is a side view when the QFP leads are not soldered.

FIG. 8 is a read image diagram when a shape illumination (rectangular shape) is applied to the solder surface when the QFP leads are not soldered.

FIG. 9 is a side view when the QFP lead floats.

FIG. 10 is a read image diagram when a shape illumination (rectangle) is applied to a solder surface when a QFP lead is floated.

FIG. 11 is a configuration diagram of a conventional shape recognition device including two image capturing devices.

[Explanation of symbols]

 1 Object for Shape Recognition 2 Shape Illumination Device 3 Image Capture Device 4 Image Processing Device 5 Computer

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Satoshi Kitazaki Unoki-nu, Heno-gun, Hebei-gun Ishikawa Prefecture, No. 2 No. 98 No. 2 PIE Co., Ltd. (72) Inventor Kokichi Nakatani Unoki-nu, Unoku-cho, Kawakita-gun, Ishikawa Prefecture No. 98, No. 2 in PFU Co., Ltd. (72) Inventor, Keisuke Koyanagi, Unoki-nu, Unoki-cho, Hebei-gun, Ishikawa Prefecture No. 98, No. 2 P-Inu Co., Ltd. (72) Hiroyuki Okada, Unoke-nu, Unoku-cho, Kawakita-gun, Ishikawa Prefecture No. 98 2 PF Corporation

Claims (2)

[Claims] 1. A shape recognition target object (1) is irradiated with arbitrary shape illumination, image data of the irradiated shape recognition target object (1) is captured, and the captured image data and image data used as a determination criterion of the shape are acquired. A shape recognition method using an image obtained by irradiating a shape recognition target object with shape illumination, characterized by recognizing the shape of a shape recognition target object (1) by comparing the stored image data of an image data library. 2. A shape lighting device (2) for irradiating a shape recognition target object (1) with arbitrary shape lighting, and a shape lighting device (2).
The image capturing device (3) that captures the image data of the shape recognition target (1) illuminated by the image recognition device, and the image data of the shape recognition target (1) that is captured by the image capturing device (3) are used as a binary pattern. The image processing device (4) and the binarized pattern obtained by processing the image data of the shape recognition object (1) by the image processing device (4) are compared with the binarized pattern of the image data used as the shape determination reference. A shape recognition device using an image obtained by irradiating a shape recognition target object with shape illumination, characterized in that the shape recognition device is configured with a computer (5) for determining a shape.