Active intervention monitoring method for dangerous elevator taking behaviors of escalator
An escalator and behavioral technology, applied in image data processing, biological neural network models, instruments, etc., can solve problems such as escalator error execution, panic, injury, etc.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0058] Example 1: Preliminary screening of infrared thermal imaging
[0059] Infrared thermal imaging is used to monitor whether there are living organisms on the steps. If the temperature of the imaging surface is greater than 35 degrees, it indicates that there are living organisms. Then turn on the video image input monitoring. If it is not living organisms, such as non-living objects such as crimping lines of boxes will be excluded.
Embodiment 2
[0060] Embodiment 2: Child identification model construction
[0061] Described convolutional neural network comprises convolutional layer, pooling layer, activation layer, fully connected layer, and adopts cross-entropy loss function to train, training process adopts stochastic gradient descent (SGD) to optimize, forms stable child behavior recognition model ;
[0062] Among them, the cross entropy loss function:
[0063]
[0064] the y i Indicates the mark of sample i, the positive class is 1, and the negative class is 0; p i Indicates the probability that sample i is predicted to be positive; N is the number of input samples.
Embodiment 3
[0065] Example 3: Judgment of Safe Behavior
[0066] Convert the preprocessed input image into HSV color space, the model space includes: hue (H), saturation (S), and brightness (V); use the color range of the yellow warning line on the step, set the threshold for two Values are used for target positioning. The yellow HSV reference ranges are as follows: [16,35], [160,255], [50,255]. The valued result is extracted to obtain the target area (ROI) of the warning line.
[0067] The above algorithm is embedded into the NVIDIA Jetson system to realize efficient offline monitoring, and the abnormal monitoring signal is input into the escalator control system for emergency control.
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


