JPS58165175A - Object detection - Google Patents

Abstract

PURPOSE:To detect quickly the shape and the position of an object with a high precision, by making only the shade image of the object incident to a photographing device as a black-level picture signal and making only the direct reflected light incident to this device as a white-level picture signal. CONSTITUTION:A substrate 2 which an object 1 to be detected is put on is illuminated with a certain brightness from the direction, which is inclined left at 45 deg. from the perpendicular by a light source 3. An image pickup camera 4 is arranged in the direction inclined right at 45 deg. from the perpendicular. The light source 3 is so set that images of the object 1 and the substrate 2 are taken as white-level pictures having the same gradation. Only the shade part of the object is taken out as the black-level picture signal from the image pickup camera 4. Thus, the position and the shape of the object are detected with a high precision only by the comparison of two densities of logical ''1'' and ''0''.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an object detection method for detecting the position of an object in its shape.

Recently, in the field of consumer equipment such as televisions and tape recorders, for example, 3- (length) x 2M (width) x 1
．． Resistance elements in the shape of Sim (height) have been developed, and by mounting such extremely small resistance elements on printed circuit boards, devices can be made smaller! 6X#lJ has been applied.

Incidentally, in this case, the resistor elements and the like are automatically mounted on the printed circuit board by various automatic mounting devices tK, thereby increasing mass production efficiency.

However, in an inspection process for inspecting whether or not the position of a resistive element or the like mounted on a printed circuit board is normal, position detection is still performed using conventional image fiber technology.

In other words, it is mounted on a printed circuit board and the nine parts of the image are displayed in multiple levels.
This iii signal is extracted as the IikIgII signal, and this iii* signal is temporarily stored in a memory locking device, and is processed to detect the 11 electronic items of the meal by one-time detection processing such as micro-extraction, or by comparison once. There is. For this reason, it takes a long time to detect the object to be inspected, and the reflectance of the object to be inspected and its background are the same, and the density gradation at the image signal level is the same or very similar. In this case, it becomes impossible to recognize the object to be inspected itself, resulting in poor inspection accuracy and the time required for the inspection process is longer than other processes.

The present invention overcomes the above-mentioned drawbacks of the prior art,
The purpose of this is to detect only the shape of the object to be detected in n# range and accurately, even if the object to be detected and its background are PLed as images with the same -# degrees. The key is to combine object inspection methods.

In order to keep this in mind, the present invention has developed a light source so that only the shadow of the object to be detected enters the #I image signal at a black level, and only the reflected light from the neck or the first class enters the stirrup as a 1m image signal at a white level. Photo 11
1! l! Determining the position II relationship with the pupil and the brightness of the light source,
t1 & image signal of the object to be detected at the analog signal stage
It is a duck that is converted into a value and the shape and position of the object to be detected are detected based on this dichotomized signal.

Hereinafter, the present invention will be explained in detail using the drawings.

Then, illumination fi' with a constant brightness is generated by light source 3 from the direction of 45 & on the left side of the figure, and 4 on the right side of the figure
8jI! It is configured such that an image of the object 1 including the substrate 2 is captured by the imaging camera 4 disposed in the direction.

When the light source I3 is powered to such an extent #brJIIf;tc, a shadow as shown by the false Kulq line of the imaging camera 4 of the object 1 is formed. Therefore, the sharpness of the light source 3 is set to be high so that the image captured by the imaging camera 4 due to the class II reflected light from the voice body 1 and the board 2 is suppressed as a 1iiiIII image with the same gradation and white level. do. Then, in order to suppress only the shadow of the object 1 as an image with a black level, 11D outputs a binary image signal from the output of the imaging camera 4 in which the shadow of the object 1 is at a black level and the other parts are at a white level. Obtainable. Then, the image pickup camera 4 is set to the base [j! 1 and the object 10 am causes a kind of halation state and is deposited as a white level image signal,
K is set so that only the shaded portion of the object 1 is extracted as a swallow-level image signal. Then, since the concentrations of substrate 2 and object 1 are the same, object 1 is shown in Figure 2 (Fig. 2) corresponding to its shape.
a)5. At a predetermined position @l as a shaded area as shown in (b)
It becomes possible to understand Cfft with good accuracy.

As a result, it is possible to detect an object 10 with high accuracy at a high speed and to detect its shape by simply comparing two 111 degrees of logic 11# cars without performing conventional feature extraction processing. In the case of digital cameras, the image signal itself has already been binarized, so the high-resolution digital image signal that was previously required is
This eliminates the need for a converter, which is extremely effective in reducing the number of crucians and lowering costs.
) The first advantage in manufacturing is that it is not necessary to pay much attention to negative effects such as external SH sound during signal amplification and transmission.

Incidentally, in the above-mentioned example, the object to be detected is detected by its shadow, but this is the opposite.

Only the Ii reflected light from the object 1 may be recognized as a white level image signal. That is, as shown in FIG. 3, the light f1g is directed in the direction Kj2fL to the right 451 with respect to the perpendicular line of the substrate 2, and the imaging camera 4 is directed upward in the vertical direction of the object 1. Then, only the direct reflected light from the object 1 enters the camera 4, and the reflected light from other parts is dispersed in the horizontal direction of the figure and does not enter the camera 4. Due to this K, the part of the object 1 is detected as an image with a white level, and the other part is detected as an image with a heat level. As a result, it is possible to obtain the same effect as the pongee example shown in FIG.

In addition, in principle, the duck according to the actual fiber example in Fig. 1 is based on the condition that the object to be detected is a book protruding on a plane, and -1 third
The implementation method shown in the figure requires that the surface of the object 10 has a certain curved surface, but objects that do not protrude on a plane but have a curved surface and objects that do not have a curved surface but protrude on a plane coexist. In such a case, if power is distributed to the imaging camera at two locations, one in the vertical direction and the other in the 451 direction, and these are used in a time-sharing manner, the shapes and positions of objects of various shapes can be quickly detected.

As is clear from the above description, in the present invention, when detecting the shape and position of an object, only the shadow part of the object is a 1jii111 signal with a black level, and only the directly reflected light is an image signal with a white level. The brightness of the light source and the relationship between the light source and the imaging device were determined so that the brightness of the light source could be suppressed by the imaging device, and the imaging signal of the object to be detected was binarized at the analog signal stage, so that the object to be detected and its background were the same. - Even with wave intensity, the position and shape of the object to be detected can be detected quickly and accurately by a simple one-time comparison, which is why it is used for position inspection of various parts mounted on printed circuit boards. The time required for the process can be greatly reduced, and at the same time, one
Solder grains protruding from the surface can also be detected.

Furthermore, the MuD conversion i! j becomes praise, etc.
There are also extremely useful effects in terms of device configuration and manufacturing.

[Brief explanation of drawings]

FIG. 1 is a diagram showing one embodiment of the present invention, FIG. 2 is a diagram showing an example of a shadow image, and FIG. 3 is a diagram showing another example of the present invention. 1...Object to be detected, hit...board, 3...
Light source, 4...imaging camera. Keishi 1 person Link Seed System Co., Ltd. Agent Masa Yamakawa 1 Itsuki (and 1 other person) 2

Claims (1)

[Claims] As an Ikl Rinka issue where only the shadow of the object to be clamped is at a black level, Kanryo established the potential relationship between the light source and the imaging device W and the light source so that only the directly reflected light enters the imaging device as a white level jl image signal. The brightness of the object to be detected is determined, the image signal of the object to be detected is binarized at the analog signal stage, and the shape and head of the object to be detected are detected based on this binary signal. Method.