TW202029123A - Building material image recognition and analysis system and method capable of calculating the actual size of the image feature for a building material - Google Patents

Abstract

A building material image recognition and analysis system and method thereof for measuring the actual size of a building material includes: a photography unit for shooting the building material to obtain an image of the building material; and an image recognition unit for receiving the building material, and then uses at least one image feature in the building material image to obtain the actual size of an image feature. According to the ratio of the image feature to the building material in the building material image, the actual size of the building material can be calculated based on the actual size of the image feature and the ratio. In this way, a size-correction member is not required to obtain the installation space of building material in the process of obtaining the actual size of building material.

Description

Building material image recognition and analysis system and method

The present invention is an image recognition technology, in particular to a building material image recognition and analysis system and method.

At present, before making building materials, it is necessary to obtain the space size of the place where the building materials are to be installed, and then make matching building materials according to the aforementioned space size. However, the current way to obtain the space size of building materials is that the personnel meet the needs. The actual measurement is carried out at the installation site, but when the installation site is far away, the use of manpower for on-site measurement will derive traffic and labor costs, resulting in high costs for building materials installation.

Therefore, in order to overcome the above-mentioned problems, a Republic of China Invention Patent No. I509213 discloses a "method for obtaining door (window) installation space dimensions by using remote images", which mainly uses a dimension correction element to install a building material. On one side of the space edge, an image including the door (window) installation space and the size correction component is captured by an image capturing device; the image is then sent to an image measurement system; and the image measurement system is used Analyze the image to find out the relationship between the door (window) installation space and an image ratio of the size correction component, and calculate the size of the door (window) installation space based on the image ratio relationship and the actual size of the size correction component Actual size; but this conventional technology still has the following shortcomings in use:

1. Because this conventional technology must place a size correction component before acquiring images, it is found that measurement difficulties are often caused by the lack of a suitable size correction component. Disturb.

2. This conventional technique cannot identify the color of the building materials in the image, so the color of the building materials produced often has different colors.

Third, in addition, the conventional technology only notes the installation location when shooting, so it is often troublesome that the installation location is wrong.

The main purpose of the present invention is to provide a building material installation space that can be obtained without a size correction component.

According to the above objective, the present invention provides a building material image recognition and analysis system for measuring the actual size of a building material, which includes: a photographing unit for photographing the building material to obtain a building material image; and an image recognition unit , Receiving the building material image, and then using at least one image feature in the building material image to obtain the actual size of an image feature, and calculating the actual size of the image feature and the ratio according to the ratio of the image feature to the building material in the building material image The actual size of the building material; in this way, in the process of obtaining the actual size of the building material, the installation space of the building material can be obtained without the size correction parts.

According to the above objective, the present invention provides a method for identifying and analyzing building materials images. The process is as follows: acquiring a building material image, identifying at least one image feature in the building material image, and obtaining the actual size of an image feature; and acquiring the image feature and the building material A ratio of a building material in the image, and the actual size of the building material is calculated based on the ratio and the actual size of the image feature.

Another object of the present invention is to further understand the correct color code of building materials, which is mainly to add a color recognition module to the image recognition unit, and the color recognition module can be used together or before and after the image recognition unit recognizes the characteristics of the building materials. Identify the color of building materials in building materials images, Then, it is compared with a standard color code stored in the building material database to provide the correct color code information for the building material, thereby avoiding the production of building materials with different colors.

Another main purpose of the present invention is to provide a positioning coordinate value of a building material installation space, so as to not only accurately obtain the location where the building material is installed, but also reduce the probability of incorrect installation of doors and windows.

(10)‧‧‧Filming Unit

(20)‧‧‧Image recognition unit

(21)‧‧‧Structural feature recognition module

(22)‧‧‧Color recognition module

(23)‧‧‧Wireless communication module

(30)‧‧‧Building Materials Database

(50)‧‧‧Positioning unit

(60)‧‧‧Spot Projection Unit

(70)‧‧‧Cloud

(100)‧‧‧Handheld device

The first figure is one of the system diagrams of the present invention.

The second figure is the second schematic diagram of the system of the present invention.

The third figure is a schematic diagram of the steps of the method of the present invention.

Please refer to the first figure, which is a preferred embodiment of a building material image recognition and analysis system of the present invention, which mainly includes: a shooting unit (10), an image recognition unit (20), and a positioning unit (50); The photographing unit (10) is used for photographing a building material to obtain a building material image: and the image recognition unit (20) receives the building material image, and recognizes at least one image feature in the building material image, and then obtains a building material image by the image feature The actual size of the image feature.

In this embodiment, the image recognition unit (20) is provided with a structural feature recognition module (21), and the structural feature recognition module (21) is used to recognize the appearance or structural features of the building materials in the image feature, and The image feature to be obtained by the aforementioned identification can be compared with data stored in a building material database (30) to obtain the actual size of the image feature, and then obtain a ratio of the image feature to the building material in the building material image, and use the The actual size of the building material is calculated by the ratio and the actual size of the image feature; by this, in the process of obtaining the actual size of the building material, the building can be obtained without the size correction parts The installation space of the material.

Also in this embodiment, the building materials database (30) contains the structural features and dimensions of each building material model in the domestic and foreign building materials industry, and the building materials include masonry, glass, ceramic tiles, doors and windows, building boards, etc. , And not limited to the foregoing, the door and window include the structural features and dimensions of the outer frame, inner sash, sealing cover, and lock, and the size of the inner sash has at least the width of an inner vertical frame and an inner The groove features on the vertical fan frame, the height of an inner fan horizontal frame, and other dimensions; and the building material database (30) is also built with a standard color code, and the image recognition unit (20) is also provided with a color recognition module (22), and when the image recognition unit (20) is in the process of recognizing the building material image, the color recognition module (22) can recognize the color of the building material in the building material image together or before and after it, and then combine it with the building material data The standard color code stored in the library (30) is compared to provide the correct color code of the building material, so as to avoid the problem of color difference in the color of a building material. In this embodiment, the color discrimination module (22) It can be a color recognition software, but not limited to the foregoing.

Please refer to the second figure, the image recognition unit (20) is further provided with a wireless communication module (23), and the building material database (30) is set in a cloud (70), so that the image recognition unit (20) The data in the building material database (30) in the cloud (70) can be obtained through the wireless communication module (23); furthermore, the wireless communication module (23) can be electrically connected to a Positioning unit (50), and the positioning unit (50) can provide the positioning coordinate value of the building material together or before and after when identifying the image of the building material, so as to obtain the correct position of the building material installation, such as the location of the building material installation In order to reduce the possibility of incorrect installation in the future, the positioning unit (50) is a Bluetooth positioning, a GPS positioning, a Wifi positioning, an RFID positioning, and an infrared/laser Positioning, a ZigBee positioning or a VWB positioning, etc., are not limited to the foregoing, and any positioning unit that can provide a positioning coordinate value belongs to the application scope of the present invention.

Please refer to the second figure again. This embodiment further includes a light point projection unit (60), and the light point projection unit (60) can project multiple light points onto the building material so that the photographing unit (10) ) The acquired building material image includes the multiple light points, and the image recognition unit (20) can also select the multiple light points in the building material image as the image feature for recognition, so as to obtain the aforementioned The actual size of the image feature is continued, and then a ratio of the image feature to the building material in the image of the building material is obtained, and the actual size of the building material is calculated based on the ratio and the actual size of the image feature; and in this embodiment, the light The point projection unit (60) is an optical radar instrument (LIDAR; LiDAR for short), a laser measuring instrument, a multi-lens sensing module, a structured light sensing module, a 3d sensing module or A TOF sensing module, etc., and not limited to the foregoing, any device or module that can project light points is the application category of the present invention.

As shown in the first and second figures, in this embodiment, the photographing unit (10), the image recognition unit (20), the building material database (30), the positioning unit (50) and the light The point projection unit (60) can be combined with a handheld device (100), and the handheld device (100) includes a video camera, a smartphone with a camera lens, a smart watch with a camera lens, and a portable A computer, a tablet computer, etc., but not limited to the foregoing.

Referring to the third figure, the process of the method of the present invention according to the above system is as follows: obtain a building material image, identify at least one image feature in the building material image, obtain the actual size of an image feature; and obtain the image feature and the building material image According to a ratio of Zhongyi Building Materials, the actual size of the building material is calculated based on the ratio and the actual size of the image feature; thereby, in the process of obtaining the actual size of the building material, the installation space of the building material can be obtained without a size correction component.

Please refer to the third figure again. Before acquiring the building material image, multiple light points can be projected on a building material so that the building material image contains the multiple light points, and the building material image The multiple light points are identified as the image feature to obtain the actual size of the image feature, and when the ratio of the image feature to the building material in the building material image is obtained, the actual size of the image feature and the ratio can be calculated The actual size of the building material is shown, and the aforementioned multiple light spots are composed of an optical radar instrument (LIDAR; LiDAR), a laser measuring instrument, a multi-lens sensing module, and a structured light sensing module , A 3D sensing module or a TOF sensing module is projected out. Any device or module that can project light points is the application category of the present invention.

After acquiring the building material image, the image feature can also be identified with the building material data in a building material database (30) to obtain the actual size of the image feature. The building material database (30) contains domestic and foreign materials. The structural features and dimensions of each building material model in the building materials industry. The building materials include masonry, glass, ceramic tiles, doors and windows, building boards, etc., and are not limited to the foregoing. The doors and windows include, for example: The structural features and dimensions of the frame, the inner fan, the sealing cover and the lock, and the size of the inner fan has at least the width of an inner fan frame, a groove feature on the inner fan frame, and an inner fan horizontal frame Height and other dimensions.

Furthermore, in the process of identification, the method of the present invention can also identify a building material color of the building material in the building material image, and then compare it with a standard color code stored in the building material database (30) to provide the correct In addition to the color code, and the method of the present invention can also provide a positioning coordinate value of the building material installation space, so as to obtain the correct position of the building material installation to reduce the possibility of future misinstallation; and the positioning coordinate value adopts a Bluetooth positioning and a GPS positioning , A Wifi positioning, an RFID positioning, an infrared/laser positioning, a ZigBee positioning or a VWB positioning, etc., but not limited to the foregoing, all the positioning technology that can provide a positioning coordinate value belongs to the present invention Application category. .

It can be seen from the above that the present invention has the same effect as the conventional technology, and saves labor, labor, cost and other benefits. In the process of obtaining building material installation space, the present invention does not require size at all. The calibration piece can overcome the troubles of the conventional technology, and at the same time, a color recognition module and a positioning unit are added to the present invention. In addition to providing the correct color code data of the building material to avoid the production of building materials of different colors, it can also accurately obtain the building materials The location of the installation has multiple benefits such as reducing misinstallation.

The scope of protection of the present invention shall be defined by the scope of the attached patent application. Anyone who is familiar with the art and makes any changes and modifications without departing from the spirit and scope of the present invention shall fall within the scope of application of the present invention. .

(10)‧‧‧Filming Unit

(20)‧‧‧Image recognition unit

(21)‧‧‧Structural feature recognition module

(22)‧‧‧Color recognition module

(30)‧‧‧Building Materials Database

(50)‧‧‧Positioning unit

(100)‧‧‧Handheld device

Claims (12)

A method for identifying and analyzing building materials images. The method is as follows: acquiring a building material image, identifying at least one image feature in the building material image, and acquiring the actual size of an image feature; and acquiring a ratio of the image feature to a building material in the building material image, The actual size of the building material is calculated based on the ratio and the actual size of the image feature. For example, the method for identifying and analyzing building materials as described in item 1 of the scope of patent application, wherein the color of building materials in the building materials image is identified to provide the correct color code of the building materials. For example, the method for identifying and analyzing building materials as described in item 1 of the scope of patent application further includes providing a building material positioning coordinate value. For example, the method for identifying and analyzing building materials as described in item 3 of the scope of patent application, wherein the building materials positioning coordinate value is composed of a Bluetooth positioning, a GPS positioning, a Wifi positioning, an RFID positioning, an infrared/laser positioning, and a ZigBee Provided by either positioning or a VWB positioning. For example, the method for identifying and analyzing building materials images as described in item 1 of the scope of patent application, wherein before acquiring the building materials image, it includes projecting a plurality of light points on a building material so that the building material image includes the multiple light points. As described in item 5 of the scope of patent application, the multiple light spots are composed of an optical radar instrument, a laser measuring instrument, a multi-lens sensor module, and a structured light sensor. Projected by one of the module, a 3d sensor module or a TOF sensor module. For example, the method for identifying and analyzing a building material image described in item 5 of the scope of patent application uses the multiple light points as the image feature for identification to obtain the actual size of the image feature. An image recognition and analysis system for building materials includes: a shooting unit for shooting a building material, To obtain a building material image; an image recognition unit receives the building material image, and then obtains the actual size of an image feature based on at least one image feature in the building material image, and obtains a ratio of the image feature to the building material in the building material image to The ratio and the actual size of the image feature calculate the actual size of the building material. For example, the building material image recognition analysis system described in item 8 of the scope of patent application, wherein the image recognition unit can recognize the color of the building material in the building material image to provide the correct color code of the building material. For example, the building material image recognition and analysis system described in item 8 of the scope of patent application further includes a light point projection unit, and the light point projection unit can project multiple light points onto the building material so that the photographing unit can obtain the The multiple light spots are included in the building material image. For example, the building material image recognition analysis system described in item 10 of the scope of patent application, wherein the light spot projection unit is an optical radar instrument, a laser measuring instrument, a multi-lens sensing module, and a structured light sensing module One of the group, a 3d sensing module or a TOF sensing module. For example, the building material image recognition and analysis system described in item 8 of the scope of patent application further includes a positioning unit, and the positioning unit can provide a building material positioning coordinate value.

Images