Image processing device

Abstract

Corrects images caused by defective pixels. [Solution] An image processing apparatus 200 according to an exemplary embodiment of the present disclosure may include: a target pixel defect determination unit 210 that determines whether a target pixel in the kernel is a defective pixel; a cluster defective pixel detection unit 220 that, if the target pixel is a defective pixel, detects at least one cluster defective pixel that is a defective pixel sharing a floating diffusion node with the target pixel; an offset correction determination unit 230 that determines whether to correct the target pixel and at least one cluster defective pixel based on the offset value of the target pixel; and a defective pixel correction unit 240 that corrects the target pixel and at least one cluster defective pixel based on the determination result of whether or not to perform the correction.

Description

Technical Field

[0001] The technical idea of the present disclosure relates to an apparatus for processing images.

Background Art

[0002] An image sensing apparatus is an apparatus that captures an optical image using the property of a photosensitive semiconductor material that reacts to light. With the development of industries such as automobiles, medicine, computers, and communications, there is an increasing demand for high-performance image sensing apparatuses in various fields such as smartphones, digital cameras, game devices, Internet of Things, robots, security cameras, and medical micro cameras.

[0003] An original image captured by an image sensing apparatus may include an abnormal image due to defective pixels. Since an image caused by such defective pixels causes a decrease in the quality of the image, a process of correcting the image caused by the defective pixels is necessary.

Summary of the Invention

Problems to be Solved by the Invention

[0004] A technical problem of the present disclosure is to provide an image processing apparatus that corrects pixel values of clustered defective pixels. A technical problem of the present disclosure is to provide an image processing apparatus that detects clustered defective pixels using a memory in which coordinates of defective pixels are stored.

[0005] A technical problem of the present disclosure is to provide an image processing apparatus that corrects pixel values of defective pixels using an offset for the defective pixels.

[0006] The technical challenges that this disclosure seeks to address are not limited to those mentioned above, and other technical challenges not mentioned above can be clearly understood by a person with ordinary skill in the art to which this disclosure pertains from the following description. [Means for solving the problem]

[0007] An image processing apparatus according to an exemplary embodiment of the present disclosure may include: a target pixel defect determination unit that determines whether a target pixel in a kernel is a defective pixel; a cluster defect pixel detection unit that, if the target pixel is a defective pixel, detects at least one cluster defect pixel that is a defective pixel sharing a floating diffusion node with the target pixel; an offset correction determination unit that determines whether to correct the target pixel and at least one cluster defect pixel based on the offset value of the target pixel; and a defect pixel correction unit that corrects the target pixel and at least one cluster defect pixel based on the determination result of whether or not to perform the correction.

[0008] According to one embodiment, the offset correction determination unit can determine whether or not to correct the target pixel and at least one cluster defect pixel if the number of cluster defect pixels is greater than or equal to a threshold.

[0009] According to one embodiment, the offset value can be determined based on the target pixel value of the target pixel and the first homogeneous pixel value of a first homogeneous pixel having the same color as the target pixel in the kernel.

[0010] According to one embodiment, the offset value may be the difference between the average value of the first homogeneous pixel values ​​and the target pixel value in dark conditions where the target pixel value is smaller than the first threshold pixel value, or in white conditions where the target pixel value is larger than the second threshold pixel value.

[0011] According to one embodiment, the offset value may be the difference between the median value of the first homogeneous pixel values ​​and the target pixel value in dark conditions where the target pixel value is smaller than the first threshold pixel value, or in white conditions where the target pixel value is larger than the second threshold pixel value.

[0012] According to one embodiment, the offset correction determination unit can determine to correct the target pixel and at least one cluster defect pixel based on the offset value if the target pixel value of the target pixel is greater than the third threshold pixel value and less than the fourth threshold pixel value, and the offset value is greater than the first threshold offset value and less than the second threshold offset value.

[0013] According to one embodiment, when it is determined that the defective pixel correction unit corrects the target pixel and at least one cluster defective pixel based on an offset value, the unit can correct the target pixel value of the target pixel and the cluster defective pixel value of at least one cluster defective pixel using a value obtained by subtracting a correction value calculated using the offset value and parameter values ​​for the offset value from the target pixel value of the target pixel.

[0014] According to one embodiment, the defective pixel correction unit can determine parameter values ​​based on the correlation between the first homogeneous pixel value of a first homogeneous pixel having the same color as the target pixel in the kernel, the target pixel value, the second homogeneous pixel value of a second homogeneous pixel having the same color as at least one cluster defective pixel, and the cluster defective pixel value of at least one cluster defective pixel.

[0015] According to one embodiment, if the offset correction determination unit cannot determine whether to correct the target pixel and at least one cluster defect pixel based on the offset value, it can correct the target pixel and the cluster defect pixel using the average or median of the first homogeneous pixel values ​​of first homogeneous pixels having the same color as the target pixel in the kernel.

[0016] According to one embodiment, the cluster defect pixel detection unit can detect at least one cluster defect pixel based on the coordinates of the defect pixels stored in the external memory. According to one embodiment, the external memory may be OTP memory (One Time Programmable memory).

[0017] An image processing apparatus according to an exemplary embodiment of the present disclosure may include: a target pixel defect determination unit that determines whether a target pixel in a kernel is a defective pixel; a cluster defect pixel detection unit that detects cluster defect pixels, which are defective pixels that share a floating diffusion node with the target pixel, if the target pixel is a defective pixel; an offset correction determination unit that determines whether to correct the target pixel based on a first offset value of the target pixel and whether to correct the cluster defect pixel based on a second offset value of the cluster defect pixel; and a defect pixel correction unit that corrects the target pixel and the cluster defect pixel based on the determination result of whether or not to perform the correction.

[0018] According to one embodiment, the first offset value can be determined based on the target pixel value of the target pixel and the first homogeneous pixel value of a first homogeneous pixel having the same color as the target pixel in the kernel, and the second offset value can be determined based on the cluster defect pixel value of the cluster defect pixel and the second homogeneous pixel value of a second homogeneous pixel having the same color as the cluster defect pixel in the kernel.

[0019] According to one embodiment, the first offset value is the difference between the average value of the first homogeneous pixel values ​​and the target pixel value in a first dark condition where the target pixel value is smaller than the first threshold pixel value or a first white condition where the target pixel value is larger than the second threshold pixel value, and the second offset value may be the difference between the average value of the second homogeneous pixel values ​​and the cluster defect pixel value in a second dark condition where the cluster defect pixel value is smaller than the first threshold pixel value or a second white condition where the cluster defect pixel value is larger than the second threshold pixel value.

[0020] According to one embodiment, the first offset value is the difference between the median value of the first homogeneous pixel values ​​and the target pixel value in a first dark condition where the target pixel value is smaller than the first threshold pixel value, or a first white condition where the target pixel value is larger than the second threshold pixel value, and the second offset value may be the difference between the median value of the second homogeneous pixel values ​​and the cluster defect pixel value in a second dark condition where the cluster defect pixel value is smaller than the first threshold pixel value, or a second white condition where the cluster defect pixel value is larger than the second threshold pixel value.

[0021] According to one embodiment, the offset correction determination unit can determine that if the target pixel value of the target pixel and the cluster defect pixel value of the cluster defect pixel are greater than the third threshold pixel value and less than the fourth threshold pixel value, and the first offset value and the second offset value are greater than the first threshold offset value and less than the second threshold offset value, then the target pixel may be corrected based on the first offset value and the cluster defect pixel may be corrected based on the second offset value.

[0022] According to one embodiment, when the defective pixel correction unit is determined to correct the target pixel based on the first offset value and correct the cluster defective pixel based on the second offset value, the target pixel value is corrected using the first value obtained by subtracting the first correction value calculated using the first offset value and the parameter value from the target pixel value of the target pixel, and the second value obtained by subtracting the second correction value calculated using the second offset value and the parameter value from the cluster defective pixel value of the cluster defective pixel is used to correct the cluster defective pixel value.

[0023] According to one embodiment, the defective pixel correction unit can determine the parameter value based on the correlation relationship among the first same-kind pixel value of the first same-kind pixel having the same color as the target pixel in the kernel, the second same-kind pixel value of the second same-kind pixel having the same color as the cluster defective pixel, the target pixel value, and the cluster defective pixel value.

[0024] According to one embodiment, the cluster defective pixel detection unit can detect the cluster defective pixel based on the coordinates of the defective pixel stored in the external memory.

[0025] The image processing apparatus according to an exemplary embodiment of the present disclosure includes a target pixel defect determination unit that determines whether a target pixel in a kernel is a defective pixel, a cluster pixel defect determination unit that determines whether at least a part of the cluster pixels sharing a floating diffusion node with the target pixel is a defective pixel, an offset correction determination unit that determines whether to correct the cluster defective pixel and the target pixel based on the offset values for each of the cluster defective pixel and the target pixel, and a defective pixel correction unit that corrects the target pixel and the cluster defective pixel based on the determination result of whether correction can be performed. The features briefly summarized above regarding the present disclosure are exemplary aspects of the detailed description of the present disclosure to be described later, and do not limit the scope of the present disclosure.

Advantages of the Invention

[0026] An image processing apparatus according to an exemplary embodiment of the present disclosure can correct the pixel values of clustered defective pixels without incorrect correction. An image processing apparatus according to an exemplary embodiment of the present disclosure can detect clustered defective pixels using a memory in which the coordinates of the defective pixels are stored.

[0027] An image processing apparatus according to an exemplary embodiment of the present disclosure can correct the pixel values of defective pixels without incorrect correction using an offset for the defective pixels.

[0028] The effects obtained by the present disclosure are not limited to the effects mentioned above, and other effects not mentioned can be clearly understood by those with ordinary knowledge in the technical field to which the present disclosure belongs from the following description.

Brief Description of the Drawings

[0029] [Figure 1] It is a block diagram showing an imaging apparatus according to an exemplary embodiment of the present disclosure. [Figure 2] It is a diagram for explaining clustered defective pixels according to an exemplary embodiment of the present disclosure. [Figure 3] It is a flowchart showing an image processing method according to an exemplary embodiment of the present disclosure. [Figure 4] It is a flowchart showing an image processing method according to an exemplary embodiment of the present disclosure. [Figure 5] It is a flowchart showing an image processing method according to an exemplary embodiment of the present disclosure. [Figure 6] It is a flowchart showing an image processing method according to an exemplary embodiment of the present disclosure. [Figure 7]This figure illustrates an image processing method according to an exemplary embodiment of the present disclosure. [Figure 8] This figure illustrates an image processing method according to an exemplary embodiment of the present disclosure. [Figure 9] This block diagram shows an example of a computing device corresponding to the image processing device in Figure 1. [Modes for carrying out the invention]

[0030] Hereinafter, embodiments of the Disclosure will be described in detail with reference to the accompanying drawings, so as to be readily implementable by a person with ordinary skill in the art to which the Disclosure pertains. However, the Disclosure can be implemented in a variety of different forms and is not limited to the embodiments described herein.

[0031] In describing embodiments of this disclosure, detailed descriptions of known configurations or functions have been omitted if it is deemed that such descriptions would unnecessarily obscure the gist of this disclosure. Furthermore, parts of the drawings that are not related to the description of this disclosure have been omitted, and similar parts have been denoted by similar reference numerals.

[0032] In this disclosure, the term "connected," "joined," or "linked" of one component to another can include not only direct connections but also indirect connections between which other components exist. Furthermore, the term "contains" or "has" another component means, unless otherwise stated, that it may contain, rather than exclude, other components.

[0033] In this disclosure, terms such as "first," "second," etc., are used solely for the purpose of distinguishing one component from another, and do not limit the order or importance of the components unless otherwise specified. Therefore, within the scope of this disclosure, a first component in one embodiment may be referred to as a second component in another embodiment, and similarly, a second component in one embodiment may be referred to as a first component in another embodiment.

[0034] In this disclosure, components that are distinguished from each other are used to clearly describe their respective characteristics and do not necessarily imply that the components are separate. That is, multiple components may be integrated to form a single hardware or software unit, or a single component may be distributed to form multiple hardware or software units. Accordingly, such integrated or distributed embodiments are also included in the scope of this disclosure, without needing to be specifically mentioned.

[0035] In this disclosure, the components described in various embodiments are not necessarily essential components, and some may be optional components. Therefore, embodiments consisting of a subset of the components described in one embodiment are also included in the scope of this disclosure. Furthermore, embodiments that further include other components in addition to the components described in various embodiments are also included in the scope of this disclosure.

[0036] In this disclosure, the terms used herein to describe positional relationships, such as top, bottom, left, and right, are provided for explanatory purposes only, and if the drawings shown herein are viewed in reverse, the positional relationships described herein may be interpreted in reverse.

[0037] In this disclosure, each of the following phrases may include any one of the items listed together with the applicable phrase, or any possible combination thereof.

[0038] Hereinafter, exemplary embodiments of this disclosure will be specifically described with reference to Figures 1 to 9. Figure 1 is a block diagram showing an imaging apparatus according to an exemplary embodiment of the present disclosure. Figure 2 is a diagram illustrating a cluster defect pixel according to an exemplary embodiment of the present disclosure. Figure 1 will be explained below with reference to Figure 2.

[0039] Referring to Figure 1, the imaging device 10 can refer to devices such as a digital still camera for taking still images or a digital video camera for shooting videos. For example, the imaging device 10 can be implemented with a digital single-lens reflex (DSLR), a mirrorless camera, or a smartphone, but is not limited to these. The imaging device 10 may be a concept that includes an image sensor and is capable of capturing a subject and generating an image.

[0040] The imaging device 10 may include an image sensing device 100, an image processing device 200, and a memory 300.

[0041] The image sensing device 100 may be a CIS (Complementary Metal Oxide Semiconductor Image Sensor) that converts incident light into electrical signals. Although not shown in Figure 1, the image sensing device may include a lens module, pixel array, sensor driver, readout circuit, and timing controller.

[0042] The image sensing device 100 can generate an image data ID for the captured image. The image data ID may be digital data obtained by converting an analog pixel signal to digital.

[0043] The image processing device 200 can perform at least one image signal processing operation on an image data ID and generate processed image data.

[0044] For example, the image processing device 200 can reduce noise in image data IDs and perform image signal processing for image quality improvement, such as demosaicing, defective pixel correction, gamma correction, color filter array interpolation, color matrix, color correction, color enhancement, and lens distortion correction. Furthermore, the image processing device 200 can compress the image data after image signal processing for image quality improvement to generate an image file, or restore image data from the image file. The image compression format may be lossless or lossy. Examples of compression formats include JPEG (Joint Photographic Experts Group) and JPEG2000 for still images. For videos, multiple frames can be compressed according to the MPEG (Moving Picture Experts Group) standard to generate a video file.

[0045] The image processing device 200 may be a computing device mounted on a chip independent of the chip on which the image sensing device 100 is mounted, but is not limited thereto. The chip on which the image sensing device is mounted and the chip on which the image processing device 200 is mounted can communicate with each other via a predetermined interface. In one embodiment, the chip on which the image sensing device is mounted and the chip on which the image processing device 200 is mounted can be realized in a single package, for example, an MCP (Multi-Chip Package), but the scope of the present invention is not limited thereto.

[0046] The image processing device 200 may include a target pixel defect determination unit 210, a cluster defect pixel detection unit 220, an offset correction determination unit 230, and / or a defect pixel correction unit 240.

[0047] The target pixel defect determination unit 210 can determine whether a target pixel in the kernel is a defective pixel. Specifically, the target pixel defect determination unit 210 can receive memory data MD from memory 300. Memory data MD may contain information about the coordinates of the defective pixel and the offset value of the defective pixel. Therefore, the target pixel defect determination unit 210 can determine whether the coordinates of the target pixel, which is the center pixel of the current kernel, are the same as the coordinates of the defective pixel. If the coordinates of the defective pixel and the target pixel are the same, the target pixel defect determination unit 210 can determine that the target pixel is a defective pixel. However, the method for determining whether a target pixel is a defective pixel is not limited to the method described above, and various methods can be used.

[0048] The cluster defect pixel detection unit 220 can detect at least one cluster defect pixel contained within a kernel. The kernel may contain pixels that share a floating diffusion node with a target pixel, and these can each be referred to as cluster pixels. Furthermore, among the cluster pixels that share a floating diffusion node with a target pixel, those that correspond to a defective pixel can be referred to as cluster defect pixels. The cluster defect pixel detection unit 220 can receive information TI about the target pixel from the target pixel defect determination unit 210, and if it determines that the target pixel is a defective pixel, the cluster defect pixel detection unit 220 can detect at least one cluster defect pixel. For example, the cluster defect pixel detection unit 220 can compare the coordinates of the defective pixel with the coordinates of the cluster pixels and detect the cluster defect pixel corresponding to the defective pixel from among the cluster pixels. Pixels containing the target pixel and the cluster defect pixel can be referred to as a defective pixel cluster.

[0049] Referring to Figure 2, assuming the kernel size is 8x8, the kernel can contain various types of defective pixel clusters. For example, Figure 2(a) has a Bayer CFA (Color Filter Array) pattern and contains 2x2 defective pixel clusters corresponding to the Bayer CFA pattern. Also, Figure 2(b) has a Bayer CFA pattern and contains 2x4 defective pixel clusters corresponding to two Bayer patterns. Furthermore, Figure 2(c) has a quad Bayer CFA pattern and contains 2x2 defective pixel clusters corresponding to red pixels. Furthermore, Figure 2(d) has a quad Bayer CFA pattern and contains 2x2 defective pixel clusters corresponding to green pixels. Furthermore, Figure 2(e) has a quad Bayer CFA pattern and contains 2x4 defective pixel clusters corresponding to red and green pixels. Furthermore, Figure 2(f) has a quad Bayer CFA pattern and contains 2x4 defective pixel clusters corresponding to green and blue pixels. The defective pixel clusters in Figure 2 are illustrative for illustrative purposes only, and the kernel size, number of clustered defective pixels, type of defective pixel cluster, etc., according to the exemplary embodiments of this disclosure are not limited to those described above.

[0050] The cluster defect pixel detection unit 220 can transmit information CD regarding whether or not a cluster defect pixel has been detected to the offset correction determination unit 230 or the defect pixel correction unit 240. For example, if a cluster defect pixel is detected, the cluster defect pixel detection unit 220 can transmit information CD indicating that a cluster defect pixel has been detected to the offset correction determination unit 230. Also, if no cluster defect pixels are detected, the cluster defect pixel detection unit 220 can transmit information CND indicating that no cluster defect pixels were detected to the defect pixel correction unit 240.

[0051] Furthermore, if the cluster defect pixel detection unit 220 detects a number of cluster defect pixels greater than or equal to a threshold, it can transmit information CD to the offset correction determination unit 230 indicating that cluster defect pixels have been detected. Also, if the cluster defect pixel detection unit 220 detects a number of cluster defect pixels less than or equal to a threshold, it can transmit information CND to the defect pixel correction unit 240 indicating that not enough cluster defect pixels were detected in the kernel.

[0052] The offset correction determination unit 230 can determine whether or not to correct the target pixel and at least one cluster-defect pixel based on the offset value. For example, if the number of cluster-defect pixels included in the kernel is greater than or equal to a threshold, the offset correction determination unit 230 can determine whether or not to correct the target pixel and at least one cluster-defect pixel based on the offset value. Here, the offset can represent the difference between the defective pixel and the surrounding pixels of the defective pixel. For example, the offset value may be a value determined by the pixel value of the defective pixel and the pixel values ​​of the surrounding pixels of the defective pixel. Here, surrounding pixels can represent pixels of the same type that have the same color as the defective pixel. The offset value may be a value determined by the pixel value of the defective pixel and the pixel values ​​of the surrounding pixels of the defective pixel under conditions darker than a predetermined brightness. Alternatively, the offset value may be a value determined by the pixel value of the defective pixel and the pixel values ​​of the surrounding pixels under conditions brighter than a predetermined brightness. For example, the offset value may be the difference between the average value of the same type pixel values ​​of the defective pixel and the defective pixel value of the defective pixel under dark conditions where the pixel value of the defective pixel is smaller than the first threshold pixel value. Alternatively, the offset value may be the difference between the average value of the same type of pixels of the defective pixel and the defective pixel value of the defective pixel, in the white condition where the pixel value of the defective pixel is greater than the second threshold pixel value. Also, the offset value may be the difference between the median value of the same type of pixels of the defective pixel and the defective pixel value, in the dark condition. Alternatively, the offset value may be the difference between the median value of the same type of pixels of the defective pixel and the defective pixel value, in the white condition. Furthermore, the offset value may be the difference between the defective pixel value of the defective pixel and the pixel value of any one of the surrounding pixels, in either the dark or white condition.The method for calculating the offset value is not limited to those described above; various methods that can show the difference between the defective pixel and surrounding pixels can be used to calculate the offset value.

[0053] An offset value can exist for each defective pixel. For example, if the target pixel is a defective pixel, the offset value for the target pixel may be a value determined based on the target pixel value of the target pixel and the first homogeneous pixel value of a first homogeneous pixel in the kernel that has the same color as the target pixel. Similarly, the offset value for a cluster defective pixel may be a value determined based on the cluster defective pixel value of the cluster defective pixel and the second homogeneous pixel value of a second homogeneous pixel in the kernel that has the same color as the cluster defective pixel.

[0054] The offset correction determination unit 230 can decide whether to correct a defective pixel based on the offset value if the defective pixel value of the defective pixel falls within a predetermined range and the offset value for the defective pixel falls within another predetermined range. More specific details on how the offset correction determination unit 230 decides whether or not to correct a defective pixel based on the offset value will be described later.

[0055] The defective pixel correction unit 240 receives information OI from the offset correction determination unit 230 regarding whether or not to perform correction based on the offset, and can generate processed image data PID by correcting the defective pixels based on this information. Specifically, the defective pixel correction unit 240 can correct the target pixel and at least one cluster defective pixel based on the offset correction determination unit 230's decision on whether or not to perform the correction. For example, if the defective pixel correction unit 240 decides to correct the target pixel and at least one cluster defective pixel based on the offset value, it can correct the target pixel value of the target pixel and the cluster defective pixel value of at least one cluster defective pixel using a value obtained by subtracting a correction value calculated using the offset value and parameter values ​​for the offset value from the target pixel value of the target pixel. More specific details on how the defective pixel correction unit 240 corrects defective pixels will be described later.

[0056] Memory 300 can store the coordinates of defective pixels and the offset value for each defective pixel. For example, memory 300 may, but is not limited to, OTP memory (One Time Programmable memory).

[0057] Figure 3 is a flowchart showing an image processing method according to an exemplary embodiment of the present disclosure. Referring to Figure 3, an image processing method according to an exemplary embodiment of the present disclosure can acquire a pixel value, the coordinates of a defective pixel, and an offset value in step S310. For example, the image processing device can acquire a pixel value from an image sensing device and acquire the coordinates of a defective pixel and the offset value of a defective pixel from memory.

[0058] In step S320, the image processing method can determine whether or not the target pixel is a defective pixel. For example, the image processing method can determine whether or not the target pixel is a defective pixel by comparing the coordinates of the defective pixel obtained from memory with the coordinates of the target pixel, which is the center pixel of the kernel.

[0059] The image processing method can bypass step S330 if the target pixel is not a defective pixel. Specifically, if the target pixel of the current kernel is not a defective pixel, the image processing method can move the kernel's position and determine whether the new target pixel in the next kernel is a defective pixel.

[0060] The image processing method can detect a cluster of defective pixels in step S340 if the target pixel is a defective pixel. Specifically, if the target pixel is a defective pixel, the image processing method can determine whether or not a cluster of defective pixels is included in the kernel by detecting clustered defective pixels, which are defective pixels that share a floating diffusion node with the target pixel. More specific details on detecting clusters of defective pixels will be described later.

[0061] If the kernel does not contain defective pixel clusters, the image processing method can correct target pixels and cluster-defect pixels in step S350 using a pre-configured method. For example, the image processing method can correct defective pixels using the average or median of the same pixel values ​​of similar pixels having the same color as the defective pixel, but the method of correcting defective pixels is not limited to this.

[0062] If the kernel contains a cluster of defective pixels, the image processing method can determine in step S360 whether the correction requirements based on the offset of the target pixel and the cluster defective pixel are met. Specifically, if the pixel value of the target pixel falls within a predetermined range and the offset value falls within another predetermined range, the image processing method can decide to correct the target pixel and at least one cluster defective pixel based on the offset value. More specific details on deciding whether or not to correct based on the offset value will be described later.

[0063] If it is decided that the image processing method will not correct defective pixels based on the offset, in step S350, the target pixels and cluster defective pixels may be corrected by a preset method. For example, the image processing method may correct defective pixels using the average or median of the same type of pixel values ​​of similar pixels having the same color as the defective pixel, but the method of correcting defective pixels is not limited to this.

[0064] If the image processing method is determined to correct defective pixels based on an offset, in step S370, the target pixels and cluster defective pixels can be corrected based on the offset value. More specific details on correcting target pixels and cluster defective pixels based on the offset value will be described later.

[0065] Figure 4 is a flowchart showing an image processing method according to an exemplary embodiment of the present disclosure. Figure 4 will be explained below with reference to Figure 3.

[0066] The steps of the image processing method in Figure 4 may be a more specific step than step S340 in Figure 3. In an exemplary embodiment of the present disclosure, the image processing method can obtain the coordinates of a defective pixel in step S341. For example, the image processing method can obtain the coordinates of a defective pixel from memory.

[0067] In step S342, the image processing method can determine whether or not a cluster pixel contains a defective pixel. Specifically, the image processing method can determine whether or not a cluster defective pixel, which is a defective pixel, is included among the cluster pixels that share a floating diffusion node with the target pixel. For example, the image processing method can detect a cluster defective pixel by comparing the coordinates of the acquired defective pixel with the coordinates of the cluster pixel.

[0068] In step S343, the image processing method can determine whether the number of clustered defective pixels is equal to or greater than a threshold. For example, assuming that four pixels share a floating diffusion node, the image processing method can determine whether the number of clustered defective pixels is two or more. In other words, since the target pixel has been determined to be a defective pixel, if the number of clustered defective pixels is two, it can be considered that the defective pixel cluster contains three defective pixels. Furthermore, assuming that eight pixels share a floating diffusion node, the image processing method can determine whether the number of clustered defective pixels is six or more. In other words, since the target pixel has been determined to be a defective pixel, if the number of clustered defective pixels is six, it can be considered that the defective pixel cluster contains seven defective pixels. The numerical values ​​described above are merely examples, and the image processing method according to the exemplary embodiments of this disclosure is not limited thereto.

[0069] The image processing method can determine in step S344 that the defective pixel cluster is included in the kernel if the number of cluster defective pixels is greater than or equal to a threshold. Referring to Figure 3, if the image processing method determines that the defective pixel cluster is included in the kernel, it can proceed to step S360 to determine whether the correction requirements based on the offset of the target pixel and the cluster defective pixel are met.

[0070] If the number of cluster-defective pixels is not above a threshold, the image processing method can determine in step S345 that the defective pixel cluster is not included in the kernel. Referring to Figure 3, if the image processing method determines that the defective pixel cluster is not included in the kernel, it proceeds to step S350, where the target pixels and cluster-defective pixels can be corrected using a pre-configured method.

[0071] Figure 5 is a flowchart showing an image processing method according to an exemplary embodiment of the present disclosure. Figure 5 will be explained below with reference to Figure 3.

[0072] The steps of the image processing method in Figure 5 may be a more specific step than step S360 in Figure 3. Referring to Figure 5, an image processing method according to an exemplary embodiment of the present disclosure can acquire pixel values ​​and offset values ​​in step S361. Specifically, the image processing method can acquire pixel values ​​from an image sensing device and acquire offset values ​​for defective pixels from memory.

[0073] The image processing method may correct defective pixels by correcting the entire defective pixel based on the offset value of the target pixel, or by correcting each defective pixel individually based on its own offset value. For example, the image processing method can correct both the target pixel and at least one cluster defective pixel using the first offset value of the target pixel. Alternatively, the image processing method can correct the target pixel using the first offset value of the target pixel, and the first cluster defective pixel using the second offset value of the first cluster defective pixel. Furthermore, if a second cluster defective pixel is included in the kernel, the second cluster defective pixel can be corrected using the third offset value of the second cluster defective pixel.

[0074] In step S362, the image processing method can determine whether the pixel value of a defective pixel is smaller than the lower limit pixel value. For example, assuming that defective pixels in the kernel are corrected using only the offset value of the target pixel, the image processing method can determine in step S362 whether the target pixel value is smaller than the lower limit pixel value (also referred to as the third threshold pixel value).

[0075] Furthermore, assuming that each defective pixel is corrected using its respective offset value, the image processing method can determine in step S362 whether the pixel value of each defective pixel (target pixel and at least one cluster defective pixel) is smaller than the lower limit pixel value. For example, assuming that the kernel contains a first cluster defective pixel and a second cluster defective pixel, the image processing method can determine whether the target pixel value of the target pixel, the first cluster defective pixel value of the first cluster defective pixel, and the second cluster defective pixel value of the second cluster defective pixel are smaller than the lower limit pixel value.

[0076] If the image processing method determines that the pixel value of a defective pixel is smaller than the lower limit pixel value, it can determine in step S367 that the correction requirements based on the offset of the target pixel and cluster defective pixel were not met. Since the correction requirements based on the offset were not met, the image processing method can correct the target pixel and cluster defective pixel using a preset method, as shown in step S350 of Figure 3.

[0077] If the image processing method determines that the pixel value of a defective pixel is not smaller than the lower limit pixel value, it can determine in step S363 whether the pixel value of the defective pixel is larger than the upper limit pixel value. For example, assuming that defective pixels in the kernel are corrected using only the offset value of the target pixel, the image processing method can determine in step S363 whether the target pixel value is larger than the upper limit pixel value (also referred to as the fourth threshold pixel value).

[0078] Furthermore, assuming that each defective pixel is corrected using its respective offset value, the image processing method can determine in step S362 whether the pixel value of each defective pixel (target pixel and at least one cluster defective pixel) is greater than the upper limit pixel value. For example, assuming that the kernel contains a first cluster defective pixel and a second cluster defective pixel, the image processing method can determine whether the target pixel value of the target pixel, the first cluster defective pixel value of the first cluster defective pixel, and the second cluster defective pixel value of the second cluster defective pixel are greater than the upper limit pixel value.

[0079] If the image processing method determines that the pixel value of a defective pixel is greater than the upper limit pixel value, it can determine in step S367 that the correction requirements based on the offset of the target pixel and cluster defective pixel were not met. Since the correction requirements based on the offset were not met, the image processing method can correct the target pixel and cluster defective pixel using a preset method, as shown in step S350 of Figure 3.

[0080] If the image processing method determines that the pixel value of a defective pixel is not greater than the upper limit pixel value, it can determine in step S364 whether the offset value of the defective pixel is less than the lower limit offset value. For example, assuming that defective pixels in the kernel are corrected using only the offset value of the target pixel, the image processing method can determine in step S364 whether the first offset value of the target pixel is less than the lower limit offset value (also referred to as the first threshold offset value).

[0081] Furthermore, assuming that each defective pixel is corrected using its respective offset value, the image processing method can determine in step S364 whether the offset value of each defective pixel (target pixel and at least one cluster defective pixel) is smaller than the lower limit offset value. For example, assuming that the kernel contains a first cluster defective pixel and a second cluster defective pixel, it is possible to determine whether the first offset value of the target pixel is smaller than the lower limit offset value, whether the second offset value of the first cluster defective pixel is smaller than the lower limit offset value, and whether the third offset value of the second cluster defective pixel is smaller than the lower limit offset value.

[0082] If the image processing method determines that the pixel value of a defective pixel is smaller than the lower limit pixel value, it can determine in step S367 that the correction requirements based on the offset of the target pixel and cluster defective pixel were not met. Since the correction requirements based on the offset were not met, the image processing method can correct the target pixel and cluster defective pixel using a preset method, as shown in step S350 of Figure 3.

[0083] If the image processing method determines that the offset value of a defective pixel is not smaller than the lower limit pixel value, it can determine in step S365 whether the offset value of the defective pixel is larger than the upper limit offset value. For example, assuming that defective pixels in the kernel are corrected using only the offset value of the target pixel, the image processing method can determine in step S365 whether the first offset value of the target pixel is larger than the upper limit offset value (also referred to as the second threshold offset value).

[0084] Furthermore, assuming that each defective pixel is corrected using its respective offset value, the image processing method can determine in step S365 whether the offset value of each defective pixel (target pixel and at least one cluster defective pixel) is greater than the upper limit offset value. For example, assuming that the kernel contains a first cluster defective pixel and a second cluster defective pixel, it can determine whether the first offset value of the target pixel is greater than the upper limit offset value, whether the second offset value of the first cluster defective pixel is greater than the upper limit offset value, and whether the third offset value of the second cluster defective pixel is greater than the upper limit offset value.

[0085] If the image processing method determines that the offset value of a defective pixel is greater than the upper limit pixel value, it can determine in step S367 that the correction requirements based on the offset of the target pixel and cluster defective pixel have not been met. Since the correction requirements based on the offset have not been met, the image processing method can correct the target pixel and cluster defective pixel using a pre-set method, as shown in step S350 of Figure 3.

[0086] Furthermore, if the image processing method determines that the offset value is not greater than the upper limit offset value, it can determine that the correction requirements based on the offset of the target pixel and cluster defect pixel have been met. Since the correction requirements based on the offset have been met, the image processing method can correct the target pixel and cluster defect pixel based on the offset value, as shown in step S370 of Figure 3.

[0087] The order of steps S362 to S365 described above is merely an example, and the order of S362 to S365 may be changed, or they may be performed simultaneously. Steps S361 to S367 described above can be represented in pseudocode as follows:

[0088] valid1=(pxl <reg_shared_pxl_lower_bound) valid2=(pxl>reg_shared_pxl_upper_bound) valid3=(offset <reg_shared_offset_lower_bound) valid4=(offset>reg_shared_offset_upper_bound) fail_pxl_valid=(valid1|valid2) fail_offset_valid=(valid3|valid4) fail_valid=(fail_pxl_valid|fail_offset_valid)

[0089] Here, pxl, offset, reg_shared_pxl_lower_bound, reg_shared_pxl_upper_bound, reg_shared_offset_lower_bound, reg_shared_offset_lower_bound, and reg_shared_offset_upper_bound can represent the pixel value of the defective pixel, the offset value of the defective pixel, the lower limit pixel value, the upper limit pixel value, the lower limit offset value, and the upper limit offset value, respectively. fail_valid can indicate that the pixel value and offset value do not satisfy the correction requirements based on the offset. Additionally, fail_pxl_valid relates to the pixel value and indicates that the correction requirements based on the offset are not met, and fail_offset_valid relates to the offset value and indicates that the correction requirements based on the offset are not met. valid1 to valid4 can correspond to steps S362 to S365, respectively. Therefore, if fail_valid is 1, the correction requirements based on the offset were not met, and the target pixel and cluster defective pixels can be corrected in a pre-configured manner, as shown in step S350 of Figure 3. Furthermore, if fail_valid is 0, the offset-based correction requirement is met, and the target pixels and cluster-defect pixels can be corrected based on the offset value, as shown in step S370 of Figure 3.

[0090] The lower pixel value, upper pixel value, lower offset value, and upper offset value are user parameters and can be determined by user settings. Furthermore, the lower pixel value, upper pixel value, lower offset value, and upper offset value may be determined in relation to the characteristics of the image sensing device. For example, the lower pixel value and lower offset value may be as low as 0, and assuming 10-bit data is used, the upper pixel value and upper offset value may be as high as 1023.

[0091] Figure 6 is a flowchart showing an image processing method according to an exemplary embodiment of the present disclosure. Figure 7 is a diagram illustrating an image processing method according to an exemplary embodiment of the present disclosure. Figure 8 is a diagram illustrating an image processing method according to an exemplary embodiment of the present disclosure. Figure 6 will be explained below with reference to Figures 3, 7, and 8.

[0092] Figure 6 may include a step in which step S370 of Figure 3 is materialized. Referring to Figure 6, in step S371, the image processing method can acquire pixel values ​​and offset values. For example, the image processing method can acquire pixel values ​​from an image sensing device and offset values ​​from memory.

[0093] The image processing method can determine the correlation between the defective pixel value and the surrounding pixel value in step S372. For example, as shown in Figure 7, assume that there is a 5x5 kernel and that a defective pixel cluster containing four defective pixels is included in the kernel. If each pixel is represented by color and coordinates, the target pixel, which is the center pixel of the kernel, can be represented by G22, and the cluster defective pixels can be represented by R23, B32, and G33, respectively. The image processing method can use the mean, median, or the like of the surrounding pixel values ​​to determine the correlation between the defective pixel value of the defective pixel and the surrounding pixel value of the surrounding pixels. For example, the image processing method can compare the target pixel value of the target pixel (G22) with the mean pixel value or median pixel value of the first homogeneous pixels of the target pixel (G00, G02, G04, G11, G13, G20, G24, G31, G40, G42, G44), which are the surrounding pixels. Furthermore, the image processing method can compare the first cluster defect pixel value of the first cluster defect pixel (R23) with the average pixel value or median pixel value of the second homogeneous pixels (R01, R03, R21, R41, R43) of the first cluster defect pixel. The image processing method can also compare the second cluster defect pixel value of the second cluster defect pixel (B32) with the average pixel value or median pixel value of the third homogeneous pixels (B10, B12, B14, B30, B34) of the second cluster defect pixel. Furthermore, the image processing method can compare the third cluster defect pixel value of the third cluster defect pixel (G33) with the average pixel value or median pixel value of the surrounding pixels, which are the fourth homogeneous pixels (G00, G02, G04, G11, G13, G20, G24, G31, G40, G42, G44) of the third cluster defect pixel.

[0094] When comparing the pixel values ​​of identical pixels for each target pixel and cluster-defect pixels and displaying them graphically, it can be shown as in Figure 8(a) or (b). In Figures 8(a) and (b), the x-axis represents the pixel value of the defective pixel, and the y-axis can represent the pixel value, mean pixel value, median pixel value, etc., of surrounding pixels (identical pixels for each defective pixel). In other words, the correlation between the target pixel and cluster-defect pixels and surrounding pixels can be derived as a linear relationship by linear regression, as shown in Figure 8(a) or (b). However, the correlations in the exemplary embodiments of this disclosure are not limited thereto.

[0095] In step S373, the image processing method can determine parameter values ​​based on correlations. The image processing method allows you to calculate the corrected pixel values ​​using the following formula.

[0096] [Formula 1] output=pxl-(α*offset 2 (+β*offset+γ)

[0097] Here, output is the corrected pixel value, pxl is the pixel value of the defective pixel, offset is the offset value of the defective pixel, and α, β, and γ can each be parameter values ​​for the offset value.

[0098] The image processing method can determine the values ​​of α, β, and γ based on the determined correlation. For example, assuming that a linear correlation with a slope of 1 is derived as shown in Figure 8(a), the image processing method can determine α=0, β=1, and γ to a non-zero value. Also, assuming that a linear correlation (any value other than 1) is derived as shown in Figure 8(b), the image processing method can determine α=0, and β and γ to non-zero values.

[0099] In the image processing method, in step S374, a correction value can be calculated using the offset value and parameter value. In this case, the correction value is (α*offset) in Equation 1. 2 It can be shown that (+β*offset+γ). Equation 1 is shown as calculating the corrected pixel value by subtracting the correction value from the pixel value, but alternatively, the image processing method may calculate the corrected pixel value by adding the correction value to the pixel value.

[0100] The image processing method may correct the pixel value of a defective pixel using only the offset value of the target pixel, or it may correct the pixel value of each defective pixel using its respective offset value. In other words, depending on the image processing method, the offset value in Equation 1 may be fixed to the offset value of the target pixel, or it may be changed to the offset value of each defective pixel.

[0101] For example, when correcting the target pixel, the first cluster defect pixel, and the second cluster defect pixel using only the first offset value of the target pixel, the corrected pixel values ​​are as follows:

[0102] Corrected target pixel value = target pixel value - {α * (1st offset value)} 2 +β*(1st offset value)+γ} Corrected first cluster defect pixel value = first cluster defect pixel value - {α*(first offset value)} 2 +β*(1st offset value)+γ} Corrected second cluster defect pixel value = second cluster defect pixel value - {α*(first offset value)} 2 +β*(1st offset value)+γ}

[0103] In contrast, when correcting the defective pixel values ​​using the offset value of each defective pixel, the corrected pixel values ​​are as follows: Corrected target pixel value = target pixel value - {α * (1st offset value)} 2 +β*(1st offset value)+γ} Corrected first cluster defect pixel value = first cluster defect pixel value - {α*(second offset value)} 2 +β*(2nd offset value)+γ} Corrected second cluster defect pixel value = second cluster defect pixel value - {α*(third offset value)} 2 +β*(3rd offset value)+γ}

[0104] In step S375, the image processing method can correct the pixel value of a defective pixel using a correction result value obtained by subtracting a correction value from the pixel value of the defective pixel. Specifically, the image processing method can correct the pixel value of a defective pixel using the output value calculated by formula 1, thereby improving the quality of the output image.

[0105] Figure 9 is a block diagram showing an example of a computing device corresponding to the image processing device in Figure 1. Referring to Figure 9, the computing device 1000 can be shown as one embodiment of the hardware configuration for performing the operation of the image processing device 200 in Figure 1.

[0106] The computing device 1000 can be mounted on a chip independent of the chip on which the image sensing device is mounted. According to one embodiment, the chip on which the image sensing device is mounted and the chip on which the computing device 1000 is mounted can be realized in a single package, for example, an MCP (Multi-Chip Package), but the scope of the present invention is not limited thereto.

[0107] The computing device 1000 may include a processor 1010, memory 1020, an input / output interface 1030, and a communication interface 1040.

[0108] The processor 1010 can process the data and / or instructions necessary to perform the operation of the image processing device 200 as described in Figure 1.

[0109] Memory 1020 can store data and / or instructions necessary for the operation of the image processing unit 200 and can be accessed by processor 1010. For example, memory 1020 can be implemented as volatile memory (e.g., DRAM (Dynamic Random Access Memory), SRAM (Static Random Access Memory), etc.) or non-volatile memory (e.g., PROM (Programmable Read Only Memory), EPROM (Erasable PROM), EEPROM (Electrically Erasable PROM), flash memory, etc.).

[0110] In other words, the computer program for operating the image processing device 200 disclosed in this document is recorded in memory 1020 and executed and processed by processor 1010, thereby enabling the operation of the image processing device 200.

[0111] The input / output interface 1030 can provide an interface that connects an external input device (e.g., keyboard, mouse, touch panel, etc.) and / or an external output device (e.g., display) with the processor 1010, enabling the transmission and reception of data.

[0112] The communication interface 1040 is configured to send and receive various types of data with external devices (e.g., application processor, external memory, etc.) and may be a device that supports wired or wireless communication.

[0113] The above description is merely illustrative of the technical concept of this disclosure, and any person with ordinary skill in the art to which this disclosure belongs can make various modifications and variations without departing from the essential characteristics of this disclosure. Therefore, the embodiments disclosed in this disclosure are for illustrative purposes only, not to limit the technical concept of this disclosure, and the scope of the technical concept of this disclosure is not limited by such embodiments. The scope of protection of this disclosure should be interpreted in accordance with the claims set forth below, and all technical concepts within an equivalent scope should be interpreted as being included in the scope of rights of this disclosure. [Explanation of symbols]

[0114] 10. Imaging device 100 Image Sensing Devices 200 Image Processing Devices 210 Target pixel defect detection unit 220 Cluster Defect Pixel Detection Unit 230 Offset Correction Determination Unit 240 Defective Pixel Correction Unit 300 memory 1000 computing devices 1010 Processor 1020 memory 1030 Input / Output Interface 1040 Communication Interface

Claims

1. A target pixel defect determination unit that determines whether or not a target pixel in the kernel is a defective pixel, If the target pixel is the defective pixel, the cluster defective pixel detection unit detects at least one cluster defective pixel that is the defective pixel sharing a floating diffusion node with the target pixel. An offset correction determination unit that determines whether or not to correct the target pixel and the at least one cluster defect pixel based on the offset value of the target pixel, A defective pixel correction unit corrects the target pixel and the at least one cluster defective pixel based on the result of the decision on whether or not to perform the correction, An image processing device, including an image processing device.

2. The offset correction determination unit is, The image processing apparatus according to claim 1, which determines whether or not to correct the target pixel and the at least one cluster-defect pixel if the number of cluster-defect pixels is greater than or equal to a threshold.

3. The aforementioned offset value is, The image processing apparatus according to claim 1, which is determined based on the target pixel value of the target pixel and the first identical pixel value of a first identical pixel having the same color as the target pixel in the kernel.

4. The aforementioned offset value is, The image processing apparatus according to claim 3, wherein in a dark condition where the target pixel value is smaller than a first threshold pixel value, or in a white condition where the target pixel value is larger than a second threshold pixel value, the difference between the average value of the first identical pixel values ​​and the target pixel value.

5. The aforementioned offset value is, The image processing apparatus according to claim 3, wherein in a dark condition where the target pixel value is smaller than a first threshold pixel value, or in a white condition where the target pixel value is larger than a second threshold pixel value, the median value of the first identical pixel values ​​is the difference between the target pixel value and the target pixel value.

6. The offset correction determination unit is, The image processing apparatus according to claim 1, wherein if the target pixel value of the target pixel is greater than the third threshold pixel value and less than the fourth threshold pixel value, and the offset value is greater than the first threshold offset value and less than the second threshold offset value, the apparatus determines to correct the target pixel and the at least one cluster defect pixel based on the offset value.

7. The defective pixel correction unit, If it is determined that the target pixel and the at least one cluster-defect pixel should be corrected based on the offset value, the image processing apparatus according to claim 1 corrects the target pixel value and the cluster-defect pixel value of the at least one cluster-defect pixel using a value obtained by subtracting a correction value calculated using the offset value and a parameter value for the offset value from the target pixel value of the target pixel.

8. The defective pixel correction unit, The image processing apparatus according to claim 7, wherein the parameter value is determined based on the correlation between the first homogeneous pixel value of a first homogeneous pixel having the same color as the target pixel in the kernel, the target pixel value, the second homogeneous pixel value of a second homogeneous pixel having the same color as the at least one cluster-defective pixel, and the cluster-defective pixel value of the at least one cluster-defective pixel.

9. The offset correction determination unit is, If it is not determined that the target pixel and the at least one cluster-defective pixel are to be corrected based on the offset value, the image processing apparatus according to claim 1 corrects the target pixel and the cluster-defective pixel using the mean or median of the first homogeneous pixel values ​​of first homogeneous pixels having the same color as the target pixel in the kernel.

10. The cluster defect pixel detection unit, The image processing apparatus according to claim 9, which detects the at least one cluster defect pixel based on the coordinates of the defect pixel stored in an external memory.

11. The aforementioned external memory is The image processing apparatus according to claim 10, wherein the OTP memory (One Time Programmable memory) is used.

12. A target pixel defect determination unit that determines whether or not a target pixel in the kernel is a defective pixel, If the target pixel is the defective pixel, a cluster defective pixel detection unit detects a cluster defective pixel which is the defective pixel that shares a floating diffusion node with the target pixel. An offset correction determination unit that corrects the target pixel based on a first offset value of the target pixel and determines whether or not to correct the cluster defect pixel based on a second offset value of the cluster defect pixel, A defective pixel correction unit corrects the target pixels and the cluster defective pixels based on the decision result of whether or not to perform the correction, An image processing device, including an image processing device.

13. The first offset value is, Determined based on the target pixel value of the target pixel and the first identical pixel value of a first identical pixel having the same color as the target pixel in the kernel. The aforementioned second offset value is, The image processing apparatus according to claim 12, which is determined based on the cluster defect pixel value of the cluster defect pixel and the second homogeneous pixel value of a second homogeneous pixel having the same color as the cluster defect pixel in the kernel.

14. The first offset value is, In a first dark condition where the target pixel value is smaller than the first threshold pixel value, or in a first white condition where the target pixel value is larger than the second threshold pixel value, the difference between the average value of the first identical pixel values ​​and the target pixel value is: The aforementioned second offset value is, The image processing apparatus according to claim 13, wherein in a second dark condition where the cluster defect pixel value is smaller than the first threshold pixel value, or in a second white condition where the cluster defect pixel value is larger than the second threshold pixel value, the difference between the average value of the second identical pixel values ​​and the cluster defect pixel value.

15. The first offset value is, In the first dark condition where the target pixel value is smaller than the first threshold pixel value, or in the first white condition where the target pixel value is larger than the second threshold pixel value, the median value of the first identical pixel values ​​is the difference between the target pixel value and the median value of the first identical pixel values. The aforementioned second offset value is, The image processing apparatus according to claim 13, wherein in a second dark condition where the cluster defect pixel value is smaller than the first threshold pixel value, or in a second white condition where the cluster defect pixel value is larger than the second threshold pixel value, the difference between the median of the second homogeneous pixel values ​​and the cluster defect pixel value.

16. The offset correction determination unit is, The image processing apparatus according to claim 12, wherein if the target pixel value of the target pixel and the cluster defect pixel value of the cluster defect pixel are greater than the third threshold pixel value and less than the fourth threshold pixel value, and the first offset value and the second offset value are greater than the first threshold offset value and less than the second threshold offset value, the target pixel is corrected based on the first offset value and the cluster defect pixel is corrected based on the second offset value.

17. The defective pixel correction unit, If it is determined that the target pixels are corrected based on the first offset value and the cluster defect pixels are corrected based on the second offset value, the target pixel value is corrected using a first value obtained by subtracting a first correction value calculated using the first offset value and parameter value from the target pixel value of the target pixel, and the cluster defect pixel value is corrected using a second value obtained by subtracting a second correction value calculated using the second offset value and parameter value from the cluster defect pixel value of the cluster defect pixel.

18. The defective pixel correction unit, The image processing apparatus according to claim 17, wherein the parameter value is determined based on the correlation between the first homogeneous pixel value of a first homogeneous pixel having the same color as the target pixel in the kernel, the second homogeneous pixel value of a second homogeneous pixel having the same color as the cluster defect pixel, the target pixel value, and the cluster defect pixel value.

19. The cluster defect pixel detection unit, The image processing apparatus according to claim 12, which detects cluster defective pixels based on the coordinates of the defective pixels stored in an external memory.

20. A target pixel defect determination unit that determines whether or not a target pixel in the kernel is a defective pixel, A cluster pixel defect determination unit that determines whether at least some of the cluster pixels that share a target pixel and a floating diffusion node are defective pixels, An offset correction determination unit that determines whether or not to correct the cluster defective pixels and target pixels among the cluster pixels that correspond to the defective pixels based on the offset values ​​for the cluster defective pixels and the target pixels, respectively. A defective pixel correction unit corrects the target pixels and the cluster defective pixels based on the decision result of whether or not to perform the correction, An image processing device, including an image processing device.

Images