hi_isp_ae_accuracy

Description

Defines the accuracy of the exposure time and gain.

Prototype

typedef struct {     
    hi_isp_ae_accuracy_type accu_type;     
    float accuracy;     
    float offset;
} hi_isp_ae_accuracy;

Members

Considerations

• The exposure time accuracy is usually linear, and the unit is line. If the exposure time accuracy is linear and the accuracy is an integer greater than 1, the number of exposure lines computed by the AE algorithm must be an integer multiple of the accuracy plus the minimum exposure time. This function can be used to compute the long/short frame exposure time in WDR mode for some sensors. For example, if the number of exposure lines of some sensors must be 2n+1, you can set the minimum exposure time to 1 or 3 and set accuracy to 2.
• A linear relationship between the exposure time and the sensor gain is the prerequisite for AE exposure allocation. That is, when the exposure is fixed, the exposure time and sensor gain can be mutually converted to keep the image luminance unchanged. If the exposure is 4096, the allocation can be exposure time (2) x gain (2048), or exposure time (4) x gain (1024). In these two cases, the image luminance remains unchanged. If this linear relationship is not met, the image flickers when the exposure time of a line changes in a highlighted environment because the exposure time is short. Some sensors, such as Panasoni MN34220, use the 1080p 30 fps initialization sequence, the exposure time must be offset by 0.8018 lines before being linearly related to the sensor gain. Therefore, the offset variable is added. It is a float variable and its unit is line, set the offset to 0.8018 so that the AE algorithm can process the offset internally. The offset relationship does not exist between the exposure time and gain of the sensor. Therefore, you need to set this parameter to 0 in cmos.c.