A Sleeper Positioning and Counting Method Based on Multi-area Grayscale Projection in Line Array Image
A technology of gray-scale projection and linear array image, which is applied in image data processing, image analysis, image enhancement, etc., can solve the problems of poor reliability and low precision, achieve easy storage and management, high-precision mileage counting, and increase beneficial information Effect
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0069] Embodiment 1: use the number of square waves in the joint projection curve to count sleepers.
[0070] Step 1: Take the qth track two-dimensional image I, and select K detection areas {R 1 ...R K}, the value range of K is 1~10, in the specific implementation process, take K=3, such as Image 6 As shown, respectively for K detection areas {R 1 ...R K} to perform horizontal projection on the pixel values to obtain K horizontal projection curves {S 1 ... S K}, detection area R 2 The corresponding horizontal projection curve S 2 like Figure 7 shown. The heights of the K detection areas are equal, and only sleepers are included in the detection area, fasteners and rails are not included;
[0071] Step 2: For K horizontal projection curves {S 1 ... S K} for fusion to obtain the joint projection curve S′;
[0072] Step 3: Use the adaptive threshold method to binarize the joint projection curve S′ to obtain the binarized curve B, such as Figure 10 shown;
[00...
Embodiment 2
[0081] Embodiment 2: counting sleepers by using the number of crests or troughs in the first derivative of the joint projection curve.
[0082] Utilize the linear array imaging system to acquire Q two-dimensional images of tracks containing sleepers, and sequentially perform the following processing on the Q two-dimensional images of tracks:
[0083] Step 1: Take the qth track two-dimensional image I, and select K detection areas {R 1 ...R K}, K ranges from 1 to 100, in the actual implementation process, K = 3, respectively for K detection areas {R 1 ...R K} to perform horizontal projection on the pixel values to obtain K horizontal projection curves {S 1 ... S K}. The heights of the K detection areas are equal, and only sleepers are included in the detection area, fasteners and rails are not included;
[0084] Step 2: For K projection curves {S 1 ... S K} for fusion to obtain the joint projection curve S′;
[0085] Step 3: Perform first-order differential filtering...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


