Struct MotionDetector 

pub struct MotionDetector { /* private fields */ }

GPU-accelerated motion detector.

Compares successive frames to detect and quantify pixel-level motion. The first call to analyze returns a zero-motion result because there is no previous frame to compare against.

Implementations

impl MotionDetector

pub fn new( sensitivity: Sensitivity, region_count_x: u32, region_count_y: u32, ) -> Self

Create a new MotionDetector.

Arguments

• sensitivity - Detection sensitivity level.
• region_count_x - Number of sub-regions horizontally (minimum 1).
• region_count_y - Number of sub-regions vertically (minimum 1).

pub fn analyze( &mut self, frame: &[u8], width: u32, height: u32, ) -> Result<MotionAnalysis>

Analyze motion between the previous frame and the current frame.

On the first call (no previous frame), returns a MotionAnalysis with all metrics at zero. On subsequent calls the current frame is compared against the previously stored frame.

Arguments

• frame - Raw pixel data (single channel / luma expected for best results; multi-channel frames work but all bytes are compared).
• width - Frame width in pixels.
• height - Frame height in pixels.

Errors

Returns GpuError::InvalidDimensions if width * height == 0. Returns GpuError::InvalidBufferSize if a previous frame was stored with a different size.

pub fn reset(&mut self)

Reset the motion detector, clearing the stored previous frame.

pub fn region_count(&self) -> u32

Return the total number of motion analysis regions (x * y).