Scene segmentation system and method based on chromatic aberration effect

Abstract

The invention relates to a scene segmentation system and method based on a chromatic aberration effect. The invention relates to the technical field of image segmentation detection, and the method comprises the steps: collecting data, and enabling the collected input data to be an RGB image with an enhanced chromatic aberration phenomenon; extracting an edge light band by using a color extraction mode; contour drawing is carried out, and high-precision edge detection is completed. According to the invention, the input data is the RGB image with enhanced chromatic aberration phenomenon, and the extraction of the edge light band is completed by using a color extraction mode, so that the field contour drawing is completed, and the high-precision edge detection is completed. Due to the fact that a color extraction mode is adopted, the needed computing power is far smaller than that of a traditional scene segmentation mode, the response speed is greatly improved, the chromatic aberration phenomenon is enhanced through optical elements, dependence on the scene illumination environment and the image color value is reduced, and the method has remarkable advantages in scenes with insufficient illumination and similar colors.

Description

Technical Field

[0001] The present invention relates to the technical field of image segmentation detection, and is a scene segmentation system and method based on chromatic aberration effect. Background Art

[0002] When using an optical camera for edge detection, there are many problems with the traditional method of setting thresholds based on color boundaries. For example, there are false edge detections at the positions of shadow parts, missed edge detections at vignetting parts, and the inability to distinguish edges due to the similar colors of foreground objects and the background. Moreover, due to the complex method and excessive computing power requirements, the response speed is too slow. Summary of the Invention

[0003] Aiming at the deficiencies of the prior art, the present invention uses the chromatic aberration phenomenon caused by the real light propagation under natural conditions, so that a real light band appears between objects and between objects and the background. This can solve the above-mentioned problems to a certain extent. Moreover, since the segmentation method is to directly extract the real light band, the computing power requirement is reduced and the response speed is greatly improved. The present invention provides a scene segmentation system and method based on chromatic aberration effect.

[0004] The present invention provides the following technical solutions:

[0005] An optical vision recognition system based on chromatic aberration effect, the system includes: an industrial camera, an optical lens, a bracket and a support rod;

[0006] A support rod is provided with a bracket at its upper end, an industrial camera is respectively provided at both ends of the bracket, and an optical lens is also provided on the bracket, and the optical lens is arranged in front of the industrial camera lens.

[0007] Preferably, according to different application scenarios and optical environments, the optical lens is replaced with convex lenses, concave lenses, plane mirrors or prisms with different refractive indexes, different Abbe numbers and different model specifications.

[0008] Preferably, the surface of the optical lens is modified or an optical coating is pasted.

[0009] A scene segmentation method based on chromatic aberration effect, the method is implemented based on an optical vision recognition system based on chromatic aberration effect, and the method includes the following steps:

[0010] Step 1: Collect data, and the input data collected is an RGB image with enhanced chromatic aberration phenomenon;

[0011] Step 2: Use the method of color extraction to complete the extraction of the edge light band;

[0012] Step 3: Draw the contour to complete high-precision edge detection.

[0013] Preferably, step 1 is specifically as follows:

[0014] Read the image file to be processed from the specified path. If the image cannot be loaded correctly, output an error message and terminate the processing. Use the image reading function in the image processing library to read the image file.

[0015] Preferably, step 2 is specifically as follows:

[0016] Step 2.1: Convert the read image from the BGR color space to the HSV color space, and screen the color regions according to the hue, saturation, and brightness characteristics of the color. Use the color space conversion function in the image processing library for the conversion.

[0017] Step 2.2: According to the characteristics under low-light conditions, define the value range of the target color in the HSV color space, and determine the lower and upper limits of each color by setting appropriate hue, saturation, and brightness thresholds.

[0018] Step 2.3: According to the defined color range, use the color range screening function in the image processing library: create a mask for each target color, mark the pixels in the image that meet the color range as white, and those that do not meet as black.

[0019] Step 2.4: Use the edge contour finding function for each color's mask image to find the edge contours of the target color regions in the image.

[0020] Preferably, step 3 is specifically as follows:

[0021] Use the edge drawing function to draw the found edge contours on the original image, and use the image display function to display the processed image.

[0022] A scene segmentation system based on the chromatic aberration effect, the system includes

[0023] A data acquisition module, which acquires data, and the acquired input data is an RGB image with enhanced chromatic aberration phenomenon;

[0024] A color extraction module, which uses the color extraction method to complete the extraction of the edge light band;

[0025] An edge detection module, which performs contour drawing to complete high-precision edge detection.

[0026] A computer-readable storage medium, on which a computer program is stored, and the program is executed by a processor to be used for implementing a method for detecting the state of a wire rope based on vision.

[0027] A computer device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, a vision-based steel wire rope state detection method is implemented.

[0028] The present invention has the following beneficial effects:

[0029] Compared with the prior art, the present invention:

[0030] The present invention actually proposes a scene segmentation method based on the enhanced chromatic aberration effect. The input data of the present invention is an RGB image with enhanced chromatic aberration phenomenon. By using the color extraction method, the extraction of the edge light band is completed, thereby completing the contour drawing and achieving high-precision edge detection. Since the color extraction method is adopted, the required computing power is much less than that of the traditional scene segmentation method, greatly improving the response speed. Moreover, the chromatic aberration phenomenon is enhanced by using optical elements, reducing the dependence on the scene illumination environment and image color values, and having significant advantages in scenes with insufficient light and similar colors.

[0031] The method of the present invention can accurately detect the edges of blue, orange-red, and red regions in an image in a low-light environment by converting the image to the HSV color space and defining the value range of specific colors according to the characteristics under low-light conditions, providing more accurate basic data for subsequent image processing and analysis, and having high practicality and application value.

[0032] The application scenarios of the present invention are extensive, and it can be mainly applied to the extraction of passable areas and water-sky boundaries for automatic driving and navigation based on visual recognition of vehicles (such as cars, ships, aircraft, etc.); in factories, it can be applied to the detection of breakpoints in processes such as sealing and gluing, metal flaw detection, weld detection, and other fields.

[0033] The present invention has strong replaceability and replaceable components: The optical sensors used in the present invention include optical cameras, visual recognition sensors, etc. The optical elements used can be replaced with convex lenses, concave lenses, plane mirrors, and prisms with different refractive indices, different Abbe numbers, and different model specifications according to different application scenarios and optical environments, and surface modification or pasting of optical coatings can be carried out on the elements. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0035] Figure 1 It is an architecture diagram of an optical vision recognition system based on the chromatic aberration effect;

[0036] Figure 2 Side view diagram of an optical vision recognition system based on chromatic aberration effect;

[0037] Figure 3 Top view of an optical vision recognition system based on chromatic aberration effect;

[0038] Figure 4 Bottom view of an optical vision recognition system based on chromatic aberration effect

[0039] Figure 5 Bottom perspective view of an optical vision recognition system based on chromatic aberration effect;

[0040] Figure 6 Schematic diagram of a scene segmentation method based on chromatic aberration effect;

[0041] Figure 7 Optical schematic diagram of the present invention. Specific embodiments

[0042] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of them. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0043] The present invention is described in detail below in conjunction with specific embodiments. Specific Embodiment 1:

[0045] According to Figures 1-7 shown, the specific optimized technical solution adopted by the present invention to solve the above technical problems is: The present invention relates to an optical vision recognition system based on chromatic aberration effect, characterized in that: the system includes: an industrial camera, an optical lens, a bracket and a support rod;

[0046] A support rod is provided with a bracket at its upper end, and an industrial camera is respectively provided at both ends of the bracket. An optical lens is also provided on the bracket, and the optical lens is arranged in front of the industrial camera lens.

[0047] The present invention can be applied to the automatic driving, navigation, etc. of vehicles (such as cars, ships, aircraft, etc.) based on visual recognition for the extraction of passable areas and water-sky boundaries; in factories for break point detection, metal flaw detection, weld detection, etc. in processes such as sealing and gluing Specific Embodiment 2:

[0049] The difference between Embodiment 2 and Embodiment 1 of the present invention is only that:

[0050] According to different application scenarios and optical environments, the optical lens is replaced with convex lenses, concave lenses, plane mirrors or prisms with different refractive indices, different Abbe numbers, and different model specifications. The present invention has strong replaceability, and the replaceable components: the optical sensors used in the present invention include optical cameras, visual recognition sensors, etc. Specific Embodiment Three:

[0052] The difference between the third embodiment and the second embodiment of the present invention lies only in:

[0053] Modifying the surface of the optical lens or pasting an optical coating. Specific Embodiment Four:

[0055] The difference between the fourth embodiment and the third embodiment of the present invention lies only in:

[0056] A scene segmentation method based on the chromatic aberration effect, the method is implemented based on an optical vision recognition system based on the chromatic aberration effect, and the method includes the following steps:

[0057] Step 1: Collect data, and the input data collected is an RGB image with enhanced chromatic aberration phenomenon;

[0058] Step 2: Use the method of color extraction to complete the extraction of the edge light band;

[0059] Step 3: Draw the contour to complete high-precision edge detection. Specific Embodiment Five:

[0061] The difference between the fifth embodiment and the fourth embodiment of the present invention lies only in:

[0062] The specific content of Step 1 is:

[0063] Read the image file to be processed from the specified path. When the image cannot be loaded correctly, output an error message and terminate the processing, and use the image reading function in the image processing library to read the image file. Specific Embodiment Six:

[0065] The difference between the sixth embodiment and the fifth embodiment of the present invention lies only in:

[0066] The specific content of Step 2 is:

[0067] Step 2.1: Convert the read image from the BGR color space to the HSV color space, screen the color regions according to the hue, saturation, and brightness characteristics of the color, and use the color space conversion function in the image processing library for conversion;

[0068] Step 2.2: According to the characteristics under low-light conditions, define the value range of the target color in the HSV color space, and determine the lower and upper limits of each color by setting appropriate hue, saturation, and brightness thresholds.

[0069] Step 2.3: According to the defined color range, use the color range screening function in the image processing library to create a mask for each target color, mark the pixels in the image that meet the color range as white, and mark those that do not as black;

[0070] Step 2.4: Use the edge contour finding function on the mask image of each color to find the edge contours of the target color areas in the image. Specific Embodiment Seven:

[0072] The difference between Embodiment Seven and Embodiment Six of the present invention is only that:

[0073] The specific content of step 3 is as follows:

[0074] Use the edge drawing function on the original image to draw the found edge contours, and use the image display function to display the processed image. Specific Embodiment Eight:

[0076] The difference between Embodiment Eight and Embodiment Seven of the present invention is only that:

[0077] The present invention relates to a scene segmentation system based on the color difference effect, and the system includes

[0078] A data acquisition module, which acquires data, and the acquired input data is an RGB image with enhanced color difference phenomenon;

[0079] A color extraction module, which uses the color extraction method to complete the extraction of the edge light band;

[0080] An edge detection module, which performs contour drawing to complete high-precision edge detection. Specific Embodiment Nine:

[0082] The difference between Embodiment Nine and Embodiment Eight of the present invention is only that:

[0083] The present invention provides a computer-readable storage medium, on which a computer program is stored, and the program is executed by a processor to be used for implementing a scene segmentation method based on the color difference effect. Specific Embodiment Ten:

[0085] The difference between Embodiment Ten and Embodiment Nine of the present invention is only that:

[0086] The present invention provides a computer device, including a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, it implements a scene segmentation method based on the color difference effect. Specific Embodiment Eleven:

[0088] The specific application principle of the method of the present invention is as follows: The specific code is as follows:

[0089] import cv2

[0090] import numpy as np

[0091] # Read the image

[0092] image_path = 'C: / Users / huawei / Desktop / seg / ORI 11.png' # Replace with the path to your image file

[0093] image = cv2.imread(image_path)

[0094] # Check if the image is loaded correctly

[0095] if image is None:

[0096] print("Error: Could not load image.")

[0097] exit()

[0098] # Convert the BGR image to HSV image

[0099] hsv = cv2.cvtColor(image, cv2.COLOR_BGR2HSV)

[0100] # Define the ranges of blue, orange - red, and red in the HSV color space (adjusted for low - light conditions)

[0101] blue_lower = np.array([100, 45, 40]) # Hue value close to 100, representing the blue area. blue_upper = np.array([124, 255, 255]) # Higher saturation and brightness values. orange_red_lower = np.array([0, 40, 40]) # Hue value close to 0, representing the red area. orange_red_upper = np.array([25, 255, 255]) # Higher saturation and brightness values. red_lower = np.array([170, 50, 50]) # Hue value close to 180, representing the red area. red_upper = np.array([180, 255, 255]) # Higher saturation and brightness values

[0102] # Create masks for blue, orange - red, and red

[0103] blue_mask = cv2.inRange(hsv, blue_lower, blue_upper)

[0104] orange_red_mask = cv2.inRange(hsv, orange_red_lower, orange_red_upper)

[0105] red_mask = cv2.inRange(hsv, red_lower, red_upper)

[0106] # Find edge contours

[0107] # Depending on the OpenCV version, findContours may return 2 or 3 values

[0108] ret, blue_contours, hierarchy = cv2.findContours(blue_mask, cv2.RETR_TREE,

[0109] cv2.CHAIN_APPROX_SIMPLE)

[0110] ret, orange_red_contours, hierarchy = cv2.findContours(orange_red_mask, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)

[0111] ret, red_contours, hierarchy = cv2.findContours(red_mask, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)

[0112] # Draw the edges on the original image

[0113] cv2.drawContours(image, blue_contours, -1, (255, 0, 0), 2) # Draw blue edges in red

[0114] cv2.drawContours(image, orange_red_contours, -1, (0, 255, 255), 2) # Draw orange - red edges in yellow

[0115] cv2.drawContours(image, red_contours, -1, (0, 0, 255), 2) # Draw the red edges in blue

[0116] # Display the result

[0117] cv2.imshow('Edge detection', image)

[0118] cv2.waitKey(0)

[0119] cv2.destroyAllWindows()

[0120] According to the actual solution of this application: This Python code uses the OpenCV library to implement edge detection for specific color regions (blue, orange - red, and red) in an image. The specific steps are as follows:

[0121] 1. Read the image from the specified path.

[0122] 2. Convert the image from the BGR color space to the HSV color space.

[0123] 3. Define the value ranges of blue, orange - red, and red in the HSV color space.

[0124] 4. Create a mask for each color.

[0125] 5. Use the cv2.findContours function to find the edge contours of each color region.

[0126] 6. Draw these edge contours on the original image.

[0127] 7. Display the processed image.

[0128] The image edge detection method based on color features of the present invention includes the following steps:

[0129] Image reading: Read the image file to be processed from the specified path. If the image cannot be loaded correctly, an error message is output and the processing is terminated. Specifically, use the image reading function (such as cv2.imread) in the image processing library (such as OpenCV) to read the image file.

[0130] Color space conversion: Convert the read image from the BGR (blue, green, red) color space to the HSV (hue, saturation, value) color space to more conveniently screen color regions according to the hue, saturation, and value features of the color. Use the color space conversion function (such as cv2.cvtColor) in the image processing library for the conversion.

[0131] Color range definition: According to the characteristics under low-light conditions, define the value ranges of the target colors (blue, orange-red, and red) in the HSV color space. By setting appropriate hue, saturation, and brightness thresholds, determine the lower and upper limits of each color.

[0132] Mask creation: According to the defined color ranges, use the color range filtering function (such as cv2.inRange) in the image processing library to create masks for each target color, mark the pixels in the image that fall within the color range as white (255), and those that do not as black (0).

[0133] Edge contour finding: Use the edge contour finding function (such as cv2.findContours) on the mask image of each color to find the edge contours of the target color regions in the image. Depending on different OpenCV versions, the return values of this function may vary. In OpenCV 4.x versions, this function returns contour and hierarchy information.

[0134] Edge drawing and display: Use the edge drawing function (such as cv2.drawContours) to draw the found edge contours on the original image, and use the image display function (such as cv2.imshow) to display the processed image.

[0135] By converting the image to the HSV color space and defining the value ranges of specific colors according to the characteristics under low-light conditions, the present invention can accurately detect the edges of blue, orange-red, and red regions in the image in a low-light environment, providing more accurate basic data for subsequent image processing and analysis, and having high practicality and application value.

[0136] The above is only a preferred embodiment of a scene segmentation system and method based on the chromatic aberration effect. The protection scope of a scene segmentation system and method based on the chromatic aberration effect is not limited to the above embodiments. All technical solutions within this concept belong to the protection scope of the present invention. It should be noted that for those skilled in the art, several improvements and changes made without departing from the principle of the present invention should also be regarded as within the protection scope of the present invention.

Claims

1. An optical vision recognition system based on chromatic aberration effect, characterized by: The system comprises: an industrial camera, an optical lens, a bracket and a support rod; A bracket is arranged at the upper end of the support rod, an industrial camera is arranged at each end of the bracket, and an optical lens is also arranged on the bracket, and the optical lens is arranged in front of the lens of the industrial camera.

2. The system according to claim 1, characterized in that: According to different application scenarios and optical environments, the optical lenses are replaced with convex lenses, concave lenses, plane mirrors or prisms with different refractive indices, different Abbe numbers, and different models and specifications.

3. The system according to claim 2, characterized in that: Modify or adhere optical coatings on the surface of optical lenses.

4. A scene segmentation method based on chromatic aberration effect, the method is implemented based on the recognition system of claim 1, characterized in that: The method comprises the following steps: Step 1: Collect data. The collected input data is an RGB image with enhanced chromatic aberration. Step 2: Use color extraction to extract edge light bands; Step 3: Draw the contour and complete high-precision edge detection.

5. The method according to claim 4, characterized in that: The step 1 is specifically as follows: Read the image file to be processed from the specified path. If the image cannot be loaded correctly, an error message is output and processing is terminated. Use the image reading function in the image processing library to read the image file.

6. The method according to claim 5, characterized in that: The step 2 is specifically as follows: Step 2.1: Convert the read image from BGR color space to HSV color space, filter the color area according to the color hue, saturation and brightness characteristics, and use the color space conversion function in the image processing library for conversion; Step 2.2: According to the characteristics of dark light conditions, define the value range of the target color in the HSV color space, and determine the lower and upper limits of each color by setting appropriate hue, saturation, and brightness thresholds; Step 2.3: Based on the defined color range, use the color range filtering function in the image processing library: create a mask for each target color, mark the pixels in the image that match the color range as white, and mark the pixels that do not match as black; Step 2.4: Use the edge contour search function for each color mask image to find the edge contour of the target color area in the image.

7. The method according to claim 6, characterized in that: The step 3 is specifically as follows: Use the edge drawing function to draw the found edge contours on the original image, and use the image display function to display the processed image.

8. A scene segmentation system based on chromatic aberration effect, characterized by: The system comprises A data acquisition module, wherein the data acquisition module acquires data, and the acquired input data is an RGB image with enhanced chromatic aberration; A color extraction module, wherein the color extraction module uses a color extraction method to complete the extraction of edge light bands; The edge detection module performs contour drawing and completes high-precision edge detection.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the method according to claims 4 to 8.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the method of claims 4-8 is implemented.

Images