Image acquisition system

The image capture system addresses blur and brightness issues in wearable cameras by adjusting shutter speed, using action recognition, and associating previous frames, enabling high-quality, efficient, and intuitive still image capture.

JP2026088709APending Publication Date: 2026-05-29TOYOTA JIDOSHA KK

Patent Information

Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
TOYOTA JIDOSHA KK
Filing Date
2024-11-19
Publication Date
2026-05-29

AI Technical Summary

Technical Problem

Wearable cameras worn near the operator's head often result in blurred first-person perspective images due to movement, making it difficult to capture high-quality still images.

Method used

An image capture system with adjustable shutter speed, action recognition, and a light-emitting unit to record high-quality still images, using a pointing motion for instruction and cropping the image to a predetermined area, and associating previous frame images for detailed context.

Benefits of technology

Suppressed blur and improved brightness in still images, allowing intuitive recording with reduced data size and network load, enhancing image quality and efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2026088709000001_ABST
    Figure 2026088709000001_ABST
Patent Text Reader

Abstract

This system suppresses blurring caused by the worker's movements in first-person perspective images, enabling the capture of high-quality first-person perspective still images. [Solution] The image capture system comprises an image acquisition unit that records and acquires multiple frame images constituting a first-person perspective video of the worker at a set shutter speed, and an image recording unit that records a portion of the multiple frame images as a first-person perspective still image. The shutter speed when recording the first-person perspective still image is set faster than the shutter speed when recording the other frame images among the multiple frame images, excluding the frame images that constitute the first-person perspective still image.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates to an image capturing system.

Background Art

[0002] There is known a technique for acquiring a first-person perspective image of an operator using a wearable camera worn by the operator (see Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] The wearable camera is often worn near the operator's head. Therefore, due to the movement of the operator, the first-person perspective image of the operator is likely to be blurred, and it may be difficult to record a high-quality first-person perspective still image.

Means for Solving the Problems

[0005] The present disclosure can be realized in the following forms.

[0006] (1) According to one embodiment of the present disclosure, an image capture system is provided. This image capture system comprises an image acquisition unit that records and acquires a plurality of frame images constituting a first-person viewpoint moving image of a worker at a set shutter speed, and an image recording unit that records a portion of the plurality of frame images as a first-person viewpoint still image, wherein the shutter speed when recording the first-person viewpoint still image is set faster than the shutter speed when recording the other frame images of the plurality of frame images, excluding the frame images constituting the first-person viewpoint still image. According to this embodiment of the image capture system, since the shutter speed when recording the first-person viewpoint still image is set faster than the shutter speed when recording the other frame images of the plurality of frame images, excluding the frame images constituting the first-person viewpoint still image, the occurrence of blur in the recorded first-person viewpoint still image can be suppressed. Therefore, high-quality first-person viewpoint still images can be recorded. (2) The image capture system of the above configuration further comprises an action recognition unit that recognizes the operator's action to instruct the operator to take a picture, and the image recording unit may record the first-person view still image when the action recognition unit recognizes the operator's action to instruct the operator to take a picture. With this configuration of the image capture system, the image recording unit records the first-person view still image when the action recognition unit recognizes the operator's action to instruct the operator to take a picture, so it is expected that the operator will consciously suppress the movement of their own head when recording the first-person view still image. For this reason, blurring in the first-person view still image to be recorded can be suppressed more effectively. (3) The image capture system of the above form further comprises a light-emitting unit that emits a flash, and the light-emitting unit may emit a flash in conjunction with the image recording unit recording a first-person view still image. According to this form of image capture system, the light-emitting unit emits a flash in conjunction with the image recording unit recording a first-person view still image, thereby compensating for the brightness of the first-person view still image and suppressing the phenomenon in which the first-person view still image becomes dark due to a fast shutter speed setting. (4) In the image capture system of the above form, the shooting instruction action is a pointing motion by the operator, and the image recording unit may crop and record a predetermined area from the first-person view still image that includes the operator's fingertip in the pointing motion. With this form of image capture system, since the shooting instruction action is a pointing motion by the operator, a first-person view still image can be recorded by the operator's intuitive and simple actions. In addition, since the image recording unit crops and records a predetermined area from the first-person view still image that includes the operator's fingertip, a desired still image with a smaller data size than the first-person view still image can be recorded. Furthermore, the operator can simultaneously instruct the recording of a first-person view still image and the cropping of a predetermined area with a single pointing motion. (5) In the image capture system of the above form, when the image recording unit records the first-person view still image, it may record a portion of the plurality of frame images that was recorded in time before the recorded first-person view still image in association with the first-person view still image. With this form of image capture system, the image recording unit records a portion of the plurality of frame images that were recorded in time before the first-person view still image in association with the first-person view still image, so that the area around the subject in time before the first-person view still image was recorded can be recorded in more detail. In addition, the amount of data when recording a portion of the plurality of frame images recorded in association with the first-person view still image is smaller than the amount of data when recording all the frame images that make up the first-person view moving image, so that the load on the communication network can be suppressed when sending and receiving data through the communication network. [Brief explanation of the drawing]

[0007] [Figure 1] This is a block diagram of the image acquisition system in the first embodiment. [Figure 2] This is an explanatory diagram illustrating a scene in which a first-person perspective still image is recorded by the image acquisition system in the first embodiment. [Figure 3]This is a flowchart of the image recording process in the first embodiment. [Figure 4] This is a flowchart of the image recording process in the second embodiment. [Modes for carrying out the invention]

[0008] A. First Embodiment: A1. Configuration of the image acquisition system 100: Figure 1 is a block diagram of the image capture system 100 in the first embodiment. Figure 2 is an explanatory diagram illustrating the recording of a first-person view still image SI by the image capture system 100 in the first embodiment. The image capture system 100 records a first-person view still image SI, which is a still image from the first-person perspective of a worker WK. The image capture system 100 is used to record still images of areas that would be blind spots for a fixed imaging system within a factory for manufacturing vehicles VC. The recorded first-person view still image SI can be used, for example, by a supervisor of worker WK to confirm the completion and quality of work performed by worker WK.

[0009] As shown in Figure 1, the image capture system 100 comprises a processor 10, a memory 20, a light-emitting unit 30, and an internal bus 40. The image capture system 100 is configured as a wearable camera that is attached to the body of a worker WK for hands-free shooting, and more specifically, as a so-called FPV (First Person View) camera that is attached to the head of a worker WK to capture first-person perspective video of the worker WK. The processor 10 functions as an image acquisition unit 12, a motion recognition unit 14, and an image recording unit 16 by executing a program pre-stored in the memory 20. Various video data acquired by the image capture system 100 is recorded in the memory 20. The light-emitting unit 30 flashes in conjunction with recording a first-person perspective still image SI. The processor 10, memory 20, and light-emitting unit 30 are connected bidirectionally via the internal bus 40.

[0010] The image acquisition system 100 can transmit a portion of the data recorded in the memory 20 to an external computer or other device via a communication network (not shown). The communication network is configured as a wireless LAN (Local Area Network), but may also be configured as a wired LAN.

[0011] The image acquisition unit 12 records and acquires multiple frame images FR that constitute a first-person perspective video of the operator WK at a set shutter speed. The motion recognition unit 14 recognizes the operator WK's shooting instruction action. The image recording unit 16, triggered by the motion recognition unit 14 recognizing the operator WK's shooting instruction action, records one of the multiple frame images FR as a first-person perspective still image SI. The details of the processing performed by the image acquisition unit 12, the motion recognition unit 14, and the image recording unit 16 will be described later.

[0012] A2. Image recording and processing: Figure 3 is a flowchart of the image recording process in the first embodiment. Image recording process refers to a series of processes performed by the image acquisition system 100 to record a first-person view still image SI. When worker WK or worker WK's supervisor instructs the execution of the image recording process, the image recording process begins.

[0013] In step S101, the image acquisition unit 12 records and acquires multiple frame images FR that constitute the first-person perspective video of worker WK at a set shutter speed. Once the image recording process starts, the image acquisition unit 12 continuously records and acquires multiple frame images FR that constitute the first-person perspective video of worker WK at a predetermined frame rate. This allows worker WK or their supervisor to review previously recorded first-person perspective video of worker WK, thereby improving the efficiency and safety of the work performed by worker WK. Furthermore, since worker WK is aware that first-person perspective video is being recorded at all times, they are expected to consciously strive to improve the efficiency and safety of their work.

[0014] First-person perspective video footage is composed of multiple frame images (FR) stitched together in sequence, resulting in a massive amount of data. Therefore, first-person perspective video footage is usually recorded in memory 20 without being transmitted over a communication network. When the first-person perspective video footage recorded in memory 20 reaches its predetermined maximum data capacity, the system is configured to delete the oldest data from memory 20.

[0015] In step S102, the motion recognition unit 14 recognizes the operator WK's shooting instruction action. A shooting instruction action is an action taken by operator WK to instruct the operator to take a picture at the timing when they want to record a first-person view still image SI. In the first embodiment, the shooting instruction action is a pointing action by operator WK. As shown in Figure 2, operator WK performs the shooting instruction action by pointing their finger at the subject they want to record.

[0016] In step S103, when the motion recognition unit 14 recognizes the operator WK's pointing motion, which is a shooting instruction motion, the image recording unit 16 records one of the multiple frame images FR as a first-person view still image SI. The shutter speed when recording the first-person view still image SI is set faster than the shutter speed when recording the other frame images FR, excluding the frame image FR that constitute the first-person view still image SI. The image shown in the lower part of Figure 2 is one of the multiple frame images FR that constitute the first-person view video of operator WK, and is also the first-person view still image SI recorded by the image recording unit 16.

[0017] The image recording unit 16 recognizes the worker WK's fingertip FT and the subject pointed to by the worker WK's fingertip FT, focuses on the subject, and records a first-person perspective still image SI. The recognition accuracy of the motion recognition unit 14 for objects such as the worker WK's fingertip FT and the subject pointed to by the worker WK's fingertip FT can be improved, for example, by repeatedly applying machine learning using AI.

[0018] The light emitting unit 30 emits a flash in accordance with the image recording unit 16 recording the first-person viewpoint still image SI. Since the shutter speed when recording the first-person viewpoint still image SI is set faster than the shutter speed when recording other frame images FR excluding the frame image FR constituting the first-person viewpoint still image SI among the plurality of frame images FR, it is considered that the first-person viewpoint still image SI is usually darker than other frame images FR. However, since the light emitting unit 30 emits a flash in accordance with the image recording unit 16 recording the first-person viewpoint still image SI, the brightness of the first-person viewpoint still image SI can be compensated.

[0019] In step S104, the image recording unit 16 cuts out a predetermined area TR including the fingertip FT of the operator WK in the pointing operation from the first-person viewpoint still image SI and records it in the memory 20. The area TR shown in FIG. 2 is a rectangular area having a predetermined number of pixels vertically and horizontally centered on the position of the fingertip FT of the operator WK.

[0020] According to the image capturing system 100 described above, since the shutter speed when recording the first-person viewpoint still image SI is set faster than the shutter speed when recording other frame images FR excluding the frame image FR constituting the first-person viewpoint still image SI among the plurality of frame images FR, the occurrence of blur in the recorded first-person viewpoint still image SI can be suppressed. Therefore, a high-quality first-person viewpoint still image SI can be recorded.

[0021] In addition, since the image recording unit 16 records the first-person viewpoint still image SI契机 when the motion recognition unit 14 recognizes the shooting instruction motion of the operator WK, it is expected that the operator WK will consciously suppress the movement of his own head when recording the first-person viewpoint still image SI. Therefore, the occurrence of blur in the recorded first-person viewpoint still image SI can be more effectively suppressed.

[0022] The light-emitting unit 30 flashes in conjunction with the image recording unit 16 recording a first-person perspective still image SI, thereby compensating for the brightness of the first-person perspective still image SI and suppressing the phenomenon of the first-person perspective still image SI becoming dark due to a fast shutter speed setting. Therefore, the supervisor of worker WK can easily determine the completion and quality of the work performed by worker WK.

[0023] Furthermore, since the shooting instruction action is a pointing motion by the operator WK, a first-person perspective still image SI can be recorded with an intuitive and simple action by the operator WK. In addition, the image recording unit 16 crops and records a predetermined area TR that includes the operator WK's fingertip FT from the first-person perspective still image SI, so a desired still image with a smaller data size than the first-person perspective still image SI can be recorded. Moreover, the operator WK can simultaneously instruct the recording of the first-person perspective still image SI and the cropping of the predetermined area TR with a single pointing motion.

[0024] B. Second Embodiment: Figure 4 is a flowchart of the image recording process in the second embodiment. The image acquisition system 100 in the second embodiment differs from the image acquisition system 100 in the first embodiment only in that step S105 is added after step S104 of the image recording process in the first embodiment. Therefore, the same reference numerals are used for processes that are the same as in the first embodiment, and their detailed descriptions are omitted.

[0025] In step S105, the image recording unit 16 records a portion of the multiple frame images FR recorded by the image acquisition unit 12, which was recorded chronologically before the recorded first-person view still image SI, in association with the first-person view still image SI. The portion of the frame images FR recorded in association with the first-person view still image SI is, for example, a series of consecutive frame images FR that are chronologically earlier than the first-person view still image SI, and is recorded as a video of about 5 to 10 seconds. The portion of the frame images FR recorded in association with the first-person view still image SI is recorded in the memory 20 together with the first-person view still image SI.

[0026] The image capture system 100 described above provides the same effects as the image capture system 100 of the first embodiment. Furthermore, the image recording unit 16 records a portion of multiple frame images FR recorded in time before the first-person view still image SI in association with the first-person view still image SI, so it can record the area around the subject in more detail in time before the first-person view still image SI was recorded. Therefore, for example, if a problem is found in the first-person view still image SI, the circumstances leading to the occurrence of that problem can be easily traced back and confirmed. In addition, the amount of data when recording a portion of multiple frame images FR recorded in association with the first-person view still image SI is smaller than the amount of data when recording all the frame images FR that constitute the first-person view moving image, so the load on the communication network can be suppressed when sending and receiving data via the communication network.

[0027] C. Other embodiments: (C1) In the first and second embodiments, the image acquisition system 100 was used in a factory for manufacturing vehicles VC, but is not limited thereto. The image acquisition system 100 may be used in inspection factories for vehicles VC, or in manufacturing or inspection factories other than those for vehicles VC, or in various locations where workers WK perform their work.

[0028] (C2) In the first and second embodiments, the wearable camera constituting the image acquisition system 100 was attached to the head of the worker WK, but is not limited thereto. The wearable camera constituting the image acquisition system 100 may be used, for example, attached to the chest of the worker WK, or hung from the neck or shoulder of the worker WK. It may also be attached to a helmet or glasses worn by the worker WK.

[0029] (C3) In the first and second embodiments, the recorded first-person view still image SI was used by the supervisor of worker WK to confirm the completion and quality of the work, but is not limited thereto. The recorded first-person view still image SI may also be used by a system capable of recognizing the completion and quality of the work to mechanically determine the completion and quality of the work.

[0030] (C4) In the first and second embodiments, the image recording unit 16 recorded one of the multiple frame images FR as a first-person view still image SI, but is not limited thereto. For example, the image recording unit 16 may record a portion of the multiple frame images FR, i.e., a series of first-person view still images SI.

[0031] (C5) In the first and second embodiments, the image capture system 100 is equipped with a light-emitting unit 30, but this may be omitted. For example, if worker WK is working in a bright space, a bright first-person perspective still image SI can be recorded without emitting a flash from the light-emitting unit 30, so the light-emitting unit 30 can be omitted.

[0032] (C6) In the first or second embodiment, the motion recognition unit 14 recognized a pointing motion by worker WK as a shooting instruction motion by worker WK, but is not limited thereto. The motion recognition unit 14 may be configured to recognize actions of worker WK using body parts other than fingers, or it may be configured to recognize voices uttered by worker WK, such as "OK".

[0033] (C7) In the first or second embodiment, the image capture system 100 is equipped with an action recognition unit 14, but this may be omitted. The image capture system 100 may be configured to record a first-person view still image SI by, for example, a physical button or a touch-operable display, and the worker WK can record a first-person view still image SI by pressing a physical button or touching the display at any time. Alternatively, the image capture system 100 may be equipped with a wearable device that can communicate with the image capture system 100, and the worker WK can record a first-person view still image SI by operating the wearable device. Furthermore, a system that can recognize the completion and quality of work performed by the worker WK may instruct the timing for recording a first-person view still image SI, thereby enabling the recording of a first-person view still image SI regardless of the worker WK's will.

[0034] (C8) In the first or second embodiment, the image recording unit 16 recorded a region TR cropped with a predetermined number of pixels centered on the position of the fingertip FT, but is not limited thereto. The image recording unit 16 may, for example, determine the region TR to be cropped so that the subject in the direction pointed to by the fingertip FT of the worker WK is located in the center of the image.

[0035] (C9) In the second embodiment, some of the multiple frame images FR recorded in association with the first-person view still image SI were a series of consecutive frame images FR that were temporally earlier than the first-person view still image SI, but are not limited to this. The multiple frame images FR recorded in association with the first-person view still image SI may, for example, be recorded with one frame image FR skipped between each other, or they may be recorded so as to span the time before and after the time when the first-person view still image SI was recorded.

[0036] This disclosure is not limited to the embodiments described above, and can be implemented in various configurations without departing from its spirit. For example, the technical features in the embodiments corresponding to the technical features in each form described in the summary of the invention can be replaced or combined as appropriate in order to solve some or all of the above-described problems, or to achieve some or all of the above-described effects. Furthermore, if a technical feature is not described as essential in this specification, it can be deleted as appropriate. [Explanation of symbols]

[0037] 100…Image acquisition system, 10…Processor, 12…Image acquisition unit, 14…Motion recognition unit, 16…Image recording unit, 20…Memory, 30…Light emission unit, 40…Internal bus, SI…First-person view still image, FR…Frame image, FT…Fingertip, TR…Area, VC…Vehicle, WK…Worker

Claims

1. It is an image capture system, An image acquisition unit that records and acquires multiple frame images that make up a first-person perspective video of a worker at a set shutter speed, The system includes an image recording unit that records a portion of the aforementioned plurality of frame images as a first-person perspective still image, An image capture system in which the shutter speed when recording the first-person perspective still image is set to be faster than the shutter speed when recording the other frame images among the plurality of frame images, excluding the frame image that constitutes the first-person perspective still image.

2. An image acquisition system according to claim 1, The unit further includes an action recognition unit that recognizes the operator's actions to give shooting instructions. The image recording unit records the first-person perspective still image when the motion recognition unit recognizes the operator's motion to instruct the operator to take a photograph, in this image capture system.

3. An image acquisition system according to claim 1, It also includes a light-emitting unit that fires a flash, The light-emitting unit is an image capture system that emits a flash in conjunction with the image recording unit recording the first-person perspective still image.

4. The image acquisition system according to claim 2, The aforementioned shooting instruction action is a pointing motion by the worker, The image recording unit is an image capture system that crops and records a predetermined area from the first-person viewpoint still image, including the fingertips of the worker during the pointing motion.

5. An image acquisition system according to claim 1, The image recording unit, when recording the first-person view still image, records a portion of the plurality of frame images that was recorded chronologically before the recorded first-person view still image, in association with the first-person view still image, in this image capture system.